Independent Impacts and Recovery Monitoring Phase 2: February-March 2016 — Synthesis Report

Synthesis Report February-March 2016 February-March Recovery Monitoring Phase 2: Recovery Monitoring Independent Impacts and Independent Impacts

Synthesis Report Independent Impacts and Recovery Monitoring Phase 2: February-March 2016

July 2016 The Asia Foundation is a nonprofit international development organization committed to improving lives across a dynamic and developing Asia. Informed by six decades of experience and deep local expertise, our programs address critical issues affecting Asia in the 21st century—governance and law, economic development, women’s empowerment, environment, and regional cooperation. In addition, our Books for Asia and professional exchange programs are among the ways we encourage Asia’s continued development as a peaceful, just, and thriving region of the world. Headquartered in San Francisco, The Asia Foundation works through a network of offices in 18 Asian countries and in Washington, DC. Working with public and private partners, the Foundation receives funding from a diverse group of bilateral and multilateral development agencies, foundations, corporations, and individuals.

Independent Impacts and Recovery Monitoring Phase 2 Synthesis Report

456 California Street, 9th Floor San Francisco, CA U.S.A. 94104 www.asiafoundation.org

The project is funded by UK aid through the UK government and the Swiss Development Cooperation. The views expressed in this report do not necessarily reflect the UK or the Swiss government’s official policies.

Cover photo: Amanda Gurung, Alok Pokharel Design: Deddeaw Laosinchai Aid and Recovery in Post-Earthquake Nepal

PREFACE

n June 2015, The Asia Foundation began a longi- and how it was determined who was eligible for a tudinal series of studies that seek to provide in- beneficiary card. Isights into the effectiveness of aid delivery and its impact on recovery in the aftermath of the disastrous The second round of research, the findings of which we earthquakes of April-May 2015 in Nepal. The studies report here, involved fieldwork in February and March track changes over time through a mix of quantita- 2016, almost one year on from the earthquakes. The tive and qualitative research methods to assess and Nepali government established a National Reconstruc- understand how local contextual factors interact tion Authority early this year and commissioned the with state and non-state provision of aid. In doing so, development of a framework for recovery and recon- the series go beyond damage assessments that have struction over the short, medium, and longer terms. tended to focus on the quantification of impacts and Around the same time, violent protests surrounding costs. They focus also on social relations, cooperation the promulgation of the new constitution, and a de- and conflict, politics and leadership, and how they, bilitating five-month blockade along the Nepal-India with current aid efforts, shape the coping strategies border, had petered out. Findings from the second of those affected. Combined with analysis of shifts in round of research thus provide a valuable snapshot government structure and policy over the course of of Nepali state performance over the course of a year the series, the studies provide valid and reliable data of political turmoil as well as a substantive baseline on the direction and magnitude of public sentiment that will allow for a future assessment of the NRA’s about state performance. They also enable a sharper performance. focus and more precise placement of recovery/reconstruction goods and services. Among the many interesting findings of the second study, the following are emblematic: Field data collection for the first study was completed two months after the quakes, with reports on findings • Okhaldhunga district needs attention; only two from in-depth fieldwork and from a large representa- percent of people in this crisis hit district received tive household survey released in parallel. At the time, food aid in this round; the Nepali government had completed a Post-Disaster • Borrowing has risen with the number of borrowers Needs Assessment and successfully organized a do- doubling and average loan size increasing by over nors’ conference to help determine the overall level of 400 percent in severely hit districts; there is great official development assistance and government funds risk of a debt trap for the most vulnerable; needed to recover from the disaster. Our first study • There is a need to focus livelihood support on affirmed the magnitude of the earthquakes’ impacts. farming which is the main source of income of Housing destruction was widespread in highly im- most people and which is recovering slower than pacted districts. In many wards in medium and lower other livelihoods; and impact districts, levels of destruction were higher than • Eighty percent of survivors in severely hit districts aggregated district level data revealed. The study also are still in contemporary shelter. noted some crucial gaps in aid distribution. Many in highly impacted wards in medium impact districts The third and fourth studies in the series are scheduled missed out. There were vast differences across dis- for September-October 2016 and March-April of 2017. tricts on how initial damage assessments were done

iii Aid and Recovery in Post-Earthquake Nepal

We thank our research partners (Democracy Resource Nepali government officials in the NRA and the Center Nepal and Interdisciplinary Analysts), our Ministry of Federal Affairs and Local Development donor partners (UK Department for International for their support. Development and Embassy of Switzerland), and

George Varughese, Ph.D. Nepal Country Representative The Asia Foundation

Patrick Barron, Ph.D. Regional Director for Conflict & Development The Asia Foundation

iv Aid and Recovery in Post-Earthquake Nepal

Acknowledgements

his report summarizes and synthesizes findings survey included Hiranya Baral, Bal Krishna Khadka, from the second round of the Independent Im- Chandra KC, and Sandeep Thapa. The data was T pacts and Recovery Monitoring (IRM) project analyzed by Sangamitra Ramachander, Anup Phayal, in post-earthquake Nepal. IRM consists of in-depth and Patrick Barron. qualitative fieldwork and a large quantitative survey conducted in earthquake affected areas. The synthesis A number of other people provided useful inputs report was written by Mark Koenig, Lena Michaels, at various stages, including in the formation of the Sasiwan Chingchit, and Patrick Barron. research questions, finalization of the sample, and analysis of the data. They include George Varughese, The qualitative work was conducted by researchers Nandita Baruah, and James Sharrock. Adrian Morel from Democracy Resource Center Nepal (DRCN), and Patthiya Tongfueng assisted with editorial support led by Sudip Pokharel and coordinated by Apurba and Deddeaw Laosinchai designed the report’s layout. Khatiwada. Analysis of the data was done by Amy Leigh Johnson, Anubhav Ajeet, Apurba Khatiwada, The Asia Foundation wishes to express its appreci- Shekhar Parajulee, and Sudip Pokharel from DRCN, ation to the informants and survey respondents in and Asia Foundation project staff, Mark Koenig, Bhaktapur, Dhading, Gorkha, Kathmandu, Lamjung, Sasiwan Chingchit, and Lena Michaels. Special thanks Nuwakot, Okhaldhunga, Ramechhap, Sindhupalchok, goes to the team of researchers for their dedication in Solukhumbu, and Syangja. the field: the lead researchers Anubhav Ajeet, Chiran Manandhar, Nayan Pokhrel, Shekhar Parajulee, The IRM project is funded by UK aid through the UK Subhash Lamichhane, and Ujjwal Prasai, and the government and the Swiss Development Cooperation, researchers Alok Pokharel, Amanda Gurung, Anurag with support from the UK Department for International Devkota, Binu Sharma, Dipti Sherchan, Ishwari Development’s Programme Partnership Arrangement Bhattarai, Janak Raj Sapkota, Prapti Adhikari, Punam with The Asia Foundation. Andy Murray (UK DFID) Limbu, Shikha Kiran Yadav, Smita Magar, and Tanka and Pia Haenni (SDC) have managed the project from Gurung. the donor side, and have provided useful inputs at every stage. The survey was implemented by a team from Interdisciplinary Analysts (IDA) led by Sudhindra The views here do not necessarily reflect the UK or Sharma. Other IDA staff who provided support to this Swiss government’s official policies.

v Aid and Recovery in Post-Earthquake Nepal

Executive Summary

his synthesis report combines and summarizes not sufficient for the cold and five percent were unable findings from the second wave of the Inde- to make repairs at all. Illness and extreme physical T pendent Impacts and Recovery Monitoring for discomfort were common hardships endured by those Accountability in Post-Earthquake Nepal (IRM), a unable to make their temporary shelter ready for the longitudinal mixed method research project designed monsoon and winter. Children, the elderly, and preg- to monitor aid impacts and patterns of recovery in nant women in temporary shelters were particularly earthquake-affected areas. The first round of research vulnerable to sickness. To avoid hardships, some in- was conducted in June 2015. Fielding of the second dividuals moved back into their damaged homes. Ill round was carried out in eleven affected districts for health was compounded by the fact that across the 36 the quantitative survey and in six districts for the wards visited in the qualitative fieldwork, 25 health qualitative component during February and March posts had been damaged by the earthquake. There 2016. Districts included those in four categories of has been little progress on repairing or rebuilding earthquake impact identified by the government’s public infrastructure that was destroyed or damaged. Post-Disaster Needs Assessment: severely hit districts Impacts on livelihoods have been widespread. While (those most affected), crisis hit districts (second high- businesses or daily wage labors were the most nega- est impact category), hit with heavy losses districts tively affected, these two groups recovered the most (third category), and a hit district (the least impacted in the first quarter of 2016. Construction workers and of those affected). unskilled laborers have found income opportunities as the need for labor for reconstruction and repairs Recovery and reconstruction have been slow due to de- has increased. In some areas, daily wages have risen lays from the government in establishing the National two-fold. However, the majority of people are farmers Reconstruction Agency (NRA) to oversee recovery ef- and they have had more difficulties recovering fully. forts. Delays were exacerbated by violent protests and Most farmers have restarted agricultural activities blockades along the Nepal-India border arising out but face remaining challenges from loss of a harvest, of dissatisfaction towards the promulgation of a new loss of livestock, displacement, and fear of landslides. constitution in September 2015. The blockade, which Tourism was still struggling to recover from the losses ended in February 2016, resulted in a severe shortage incurred by the earthquake. and price hike of fuels and other goods all over Nepal, hampering recovery and impacting aid. Fifty-four per- One year on from the earthquakes, cash and shelter cent of survey respondents reported that the fuel crisis have become higher priority needs. The proportion had led to a reduction, or even stopping, of aid flows to of people reporting cash as an immediate priority has their community. It is within this context that IRM-2 risen from 41% in IRM-1 to 69% in IRM-2. Reconstruc- presents a picture of recovery, aid effectiveness, and tion materials are now identified by over one-third of coping strategies of people in affected areas, almost respondents as a top-two priority. The need for cash one year on from the earthquakes. and building materials are greatest in the most affected severely hit districts but are also high elsewhere. Immediate needs for shelter are greatest in Nuwakot, Earthquake impacts and needs Ramechhap, and Sindhupalchowk. Food, while still an important priority, has slightly declined in importance As of March 2016, almost 80% of people in severely since IRM-1 with demand for food higher in severely hit districts, those that were most affected by the hit districts. The only districts with high immediate earthquake, were living in temporary shelters. Most food needs among the non-severely hit districts are were able to make repairs to their shelters to get them Okhaldhunga, where food aid has fallen to very low through the winter but 21% percent said repairs were levels, and Solukhumbu, which is the poorest of the vi Aid and Recovery in Post-Earthquake Nepal

surveyed districts. While food aid sharply decreased number of people taking loans and the size of loans. from IRM-1 to IRM-2, food consumption did not fall Forty-two percent of people in severely hit districts drastically. Food aid has been effectively targeted at report borrowing since the 2015 monsoon, primarily areas with higher food insecurity and appears to have for livelihoods but also for food and shelter. Borrow- reduced the need to borrow for food. Water and sani- ing is likely to further increase in the future. Sources tation were also identified as priority needs by people of borrowing have changed since IRM-1, with more on the ground. VDC and district government officials people turning to formal sources which charge lower appear to have a good understanding of local needs. rates of interest. However, the number of people bor- The number of people stating they do not need aid any rowing from moneylenders has increased and they more has fallen since IRM-1. This suggests that some continue to lend the largest amounts per borrower. people have realized that the challenges of recovering Some groups face credit constraints, either being less have been greater than they initially expected. likely to receive loans or receiving smaller amounts. Low caste people, for example, borrow the least per person and higher income households take out much Earthquake aid larger loans than others. The findings point to too risks: that some groups vitally in need of credit miss The government continued to be the largest aid out, slowing their recovery; and that those who can provider, followed by NGOs and INGOs. Overall, the access credit are unable to repay their loans leading number of people receiving aid in IRM-2 has declined to debt traps in the longer run. in almost all districts. While aid is still primarily concentrated in severely hit districts, there is a trend Other coping strategies have been less prominent. towards decreased coverage in more affected districts Remittances have increased in importance as an and increased coverage in less affected districts. income source, but do not appear to have risen in Government aid activities decreased after the last volume. Migration has been relatively low. Six percent monsoon and focused primarily on cash assistance for of households migrated after the earthquake but most temporary shelter construction and the winter. NGOs have returned home. Asset sales have been very low, and INGOs provided more diverse range of aid but all with 6% selling some assets since the earthquake. The of these three main providers have increased their aid vast majority of asset sales (89%) have been livestock. in hit with heavy loss districts (the third category of There has been almost no sale of land, in part because impact). The trend is inconsistent across districts with people do not want to make such sales unless they need aid remaining high in Solukhumbu (95% in IRM-2) to, in part because the market for land has collapsed. whilst reducing significantly in Okhaldhunga (from 95% in IRM-1 to 58% in IRM-2). Politics and leadership Government coordination mechanisms functioned well after the earthquake but have become less active Political parties continue to play a major role in deci- since the monsoon. Lack of coordination between sion-making around aid. They are consulted by local different levels of government has meant that both officials in much the same way they are in other areas local officials and earthquake-affected people do not of local governance. Parties have played useful roles, have the information they need to plan their strategy channeling information between local government and for recovery. People do not understand whether or not citizens. However, there has been rising dissatisfaction they will be entitled to reconstruction assistance and with the role parties have played in responding to when money will arrive. Coordination between the post-earthquake needs, with 61% reporting unhappi- government and other aid providers has also become ness. A key reason for this dissatisfaction is percep- less effective and there has been extremely limited tions that the damage assessments were politicized citizen participation in targeting or monitoring aid. and declining levels of aid have also contributed. Lack of transparency and consistency on how houses are classified across a series of damage assessments There is no evidence of emerging new leadership has led to complaints and tensions and these increased challenging existing political dynamics. Ward Citizen once people realized that such classifications would Forums have played an increasing role since IRM-1, determine future access to aid. Despite these problems, but this has largely related to aid distribution with key satisfaction with most aid providers has increased decisions still taken elsewhere. There are some indi- since IRM-1. cations of changes in political preferences but most voters say they are still undecided as to who they will vote for in the next election. Coping strategies Borrowing has been the most common coping strategy. There have been large increases in both the

vii Aid and Recovery in Post-Earthquake Nepal

Social relations and conflict ( 2 ) Provide more accessible opportunities for afforda- Feelings of safety have improved since IRM-1. Where- ble credit and more cash grants to avoid debt traps as 83% of respondents in IRM-1 reported feeling • Expand soft loans programs and advertise them safe in their communities, this has increased to 97% more widely, especially in rural areas in IRM-2. There do not appear to be any changes in • Regulate interest rates of informal lenders crime rates and there has been almost no violence. While crime rates have stayed steady, some cases of ( 3 ) Adjust strategies so aid delivered fits with evolving gender-based violence, and issues with alcohol intake, needs were reported. It is unclear whether these problems • Focus on cash support and housing have increased since the earthquake, as they were also • Extend food aid present before, but these issues should be monitored closely. Perceptions of safety do not vary between men • Develop strategies to help farmers recover and women but people in temporary shelters are much • Strengthen channels so local communities can more likely to feel insecure than others. express their needs

In most communities, social relations have remained ( 4 ) Enhance communication with affected communi- strong. The early post-earthquake period saw people ties and local officials coming together to help each other respond to the dif- • Develop a communications strategy outlining ficulties they faced, and this sometimes crossed caste information on recovery programs and other identity lines. In one-quarter of the wards • Enhance formal information sharing between studied in the qualitative fieldwork, some tensions levels of government had arisen, but these have not led to open conflict or violence. Issues generally related to displacement and ( 5 ) Develop new coordination strategies resettlement, perceived discrimination or unfair treat- • Find ways to even out aid across districts ment in relief distribution, and conflict over scarce • Revisit formal coordination mechanisms, ensur- water. Several cases had a perceived caste or ethnic ing clear roles and responsibilities dimension to them. People widely feel that those of all identities have had equal access to aid and that ( 6 ) Think through targeting strategies within com- aid distribution by Village Development Committees/ munities Municipalities has been fair. However, the qualitative • Ensure targeting is context-sensitive and based fieldwork found cases of perceived unfair distribution. on clearly communicated criteria Both equal distribution of aid and more selective tar- • Expand local mediation capacity geting based on need have caused problems at times. ( 7 ) Develop coherent policy for supporting displaced persons Conclusions and recommendations • Complete geological assessments • Develop policy for permanent resettlement The report presents independent recommendations, • Clarify policy on use of public land for the dis- which are not necessarily those of the UK or Swiss placed governments: ( 8 ) Expand the recovery effort beyond physical re- ( 1 ) Focus on short and medium term improvements construction to temporary shelters • Better understand the earthquake’s impact on • Develop a strategy to provide new, or improve education, health care, and inclusion existing, temporary shelter • Increase investment in psycho-social programs, • Prioritize programs to mitigate the consequences especially for children of staying in temporary shelter

viii Aid and Recovery in Post-Earthquake Nepal

LIST OF ACRONYMS

CBS Central Bureau of Statistics CDO Chief District Officer CGI Corrugated Galvanized Iron CPN-UML Communist Party of Nepal (Unified Marxist Leninist) DDC District Development Committee DDRC District Disaster Relief Committee DLSA District Lead Support Agencies DRCN Democracy Resource Center Nepal IDA Interdisciplinary Analysts INGO International non-governmental organization IRM Independent Impacts and Recovery Monitoring for Accountability in Post-Earthquake Nepal project MJF-Nepal Madhesi Jana Adhikar Forum, Nepal NeKSAP Nepal Food Security Monitoring Program NGO Non-governmental organization NMKP Nepal Mazdoor Kizan Party NPR Nepali Rupees NRA National Reconstruction Authority PDNA Post-Disaster Needs Assessment RDC Relief Distribution Committee RPP Rastriya Prajatantra Party RPP-N Rastriya Prajatantra Party Nepal UCPN (M) Unified Communist Party of Nepal (Maoist) UN United Nations UNICEF United Nations Children’s Emergency Fund UN OCHA United Nations Office for the Coordination of Humanitarian Affairs VDC Village Development Committee WCF Ward Citizen Forum

ix Aid and Recovery in Post-Earthquake Nepal

TABLE OF CONTENTS

PREFACE III ACKNOWLEDGEMENTS V EXECUTIVE SUMMARY VI LIST OF ACRONYMS IX LIST OF FIGURES XI LIST OF TABLES XIII LIST OF CASE STUDIES XV LIST OF MAPS XV

Chapter 1. Introduction 1 1.1 Background 1 1.2 Methodology 3 1.3 Contextual changes since June 2015 6 1.4 Report structure 8

Chapter 2. Earthquake Impacts and Needs 9 2.1 Key needs 9 2.2 Shelter 12 2.4 Food 16 2.4 Livelihoods 18 2.5 Infrastructure and service delivery 22

Chapter 3. Earthquake Aid 25 3.1 Aid received since the 2015 monsoon 25 3.2 Providers of aid 32 3.3 Coordination, accountability, and transparency 34 3.4 Damage assessments and beneficiary cards 37 3.5 Satisfaction with aid providers 42 3.6 Aid to different population groups 44

Chapter 4. Coping Strategies 47 4.1 Borrowing 47 4.2 Remittances 53 4.3 Migration 55 4.4 Asset Sales 56

Chapter 5. Politics and Leadership 57 5.1 Political parties 57 5.2 Emergence of new leadership and political party preferences 60

Chapter 6. Social Relations and Conflict 61 6.1 Security and crime 61 6.2 Trust and social cohesion 63 6.3 Sense of fairness and grievances with aid providers 65 6.4. Potential sources of conflict 67

Chapter 7. Conclusions and Recommendations 69 7.1 Overview of conclusions 69 7.2 Implications and recommendations 70 x Aid and Recovery in Post-Earthquake Nepal

LIST OF FIGURES

Figure 1.1: Analytic framework 2 Figure 1.2: Effect of the crisis on access to fuel for cooking and transportation (IRM-2) 6 Figure 1.3: Share saying the fuel crisis affected the prices paid for food types (IRM-2) 7 Figure 1.4: Share saying protests over the new Constitution and the fuel crisis affected 7 aid assistance in people’s ward – by district impact (IRM-2)

Figure 2.1: Priority needs (share in top two needs) – all districts (comparison IRM-1 10 immediate needs, IRM-1 three-month needs, IRM-2 immediate needs) Figure 2.2: Where people are living now – by district impact (IRM-2) 12 Figure 2.3: Share of people reporting food as among their top two needs immediately 16 and in three months – by district (IRM-2) Figure 2.4: Share of people whose source of income was affected by the earthquake – 19 by district impact (IRM-2) Figure 2.5: Housing damage and livelihood damage – by urban/rural (IRM-2) 20 Figure 2.6: Share of affected sources of income that have improved in the past 20 three months – by source of income, all districts (IRM-2) Figure 2.7: Reconstruction of damaged infrastructure (IRM-2 qualitative research) 23 Figure 2.8: Satisfaction with public services (IRM-2) 23 Figure 2.9: Dissatisfaction with services – individual panel data (IRM-1/IRM-2 comparison) 24

Figure 3.1: Share of people receiving different types of aid – all districts 26 (IRM-1/IRM-2 comparison) Figure 3.2: Share of people receiving different types of aid – severely hit districts 27 (IRM-1/IRM-2 comparison) Figure 3.3: Quantities of cash (NPR) among those who received cash, from all sources 28 – by district impact (IRM-1/IRM-2 comparison) Figure 3.4: Share of people who have received cash – by whether or not received 29 beneficiary cards (IRM-2) Figure 3.5: Source of aid – all districts (IRM-1/IRM-2 comparison) 32 Figure 3.6: Share of people receiving government aid – by district impact 33 (IRM-1/IRM-2 comparison) Figure 3.7: Share of people receiving NGO aid – by district impact 33 (IRM-1/IRM-2 comparison) Figure 3.8: Share of people receiving INGO aid – by district impact 34 (IRM-1/IRM-2 comparison) Figure 3.9: Share of people who have received beneficiary card and whose house 38 was classified as fully damaged – by district (IRM-2) Figure 3.10: Satisfaction with aid providers – all districts, among those 42 who received aid only (IRM-1/IRM-2 comparison) Figure 3.11: Satisfaction with aid providers among those who received aid – by caste (IRM-2) 45

Figure 4.1: Share of people who have borrowed since June 2015 – by district impact 48 (IRM-1/IRM-2 comparison) Figure 4.2: Reasons for borrowing, share of those borrowing – by district impact (IRM-2) 48 Figure 4.3: Sources of borrowing among those who borrowed (IRM-1/IRM-2 comparison) 51 Figure 4.4: Monthly interest rates for different sources (IRM-1/IRM-2 comparison) 51

xi Aid and Recovery in Post-Earthquake Nepal

Figure 4.5: Sources of borrowing among those who borrowed – by gender (IRM-2) 52 Figure 4.6: Changes in remittances received from abroad (IRM-2) 54 Figure 4.7: Reasons for migration (IRM-2) 55 Figure 4.8: Share of people selling assets – by district (IRM-2) 56

Figure 5.1: Satisfaction with local political parties – by district impact (IRM-2) 58

Figure 6.1: How safe and secure do you feel now in your community? – by district 62 impact (IRM-2) Figure 6.2: Assessment of trends in social relations post-earthquake by ward respondents 63 Figure 6.3: Do you think people of every caste, religion, and ethnicity are equally 65 able to receive aid according to their needs? – by district impact, urban/rural, and gender (IRM-2) Figure 6.4: Do you think people of every caste, religion, and ethnicity are equally able 67 to receive aid according to their needs? – by ethnic group (IRM-2)

xii Aid and Recovery in Post-Earthquake Nepal

LIST OF TABLES

Table 2.1: Priority immediate needs (share in top two needs) – by district impact (IRM-2) 11 Table 2.2: Where people are living now – by district (IRM-2) 12 Table 2.3: Share of people reporting shelter as priority immediate need – by district 13 impact and district (IRM-2) Table 2.4: Were you able to make sufficient repairs to your shelter for the winter? – 14 by district impact and district, those living in temporary shelters (IRM-2) Table 2.5: Share of people reporting food as priority immediate need – by district 16 impact (IRM-2) Table 2.6: Food aid, borrowing, food consumption, and current need for food – 17 by district impact and district (IRM-1 and IRM-2) Table 2.7: Food aid, borrowing, consumption, and need for food – 18 by level of food insecurity (IRM-2) Table 2.8: Share of people whose source of income was affected by the earthquake – 19 by district impact and district (IRM-2)

Table 3.1: Share of people not receiving aid – by district impact and district 26 (IRM-1/IRM-2 comparison) Table 3.2: Amount of cash received (NPR) and share who have received cash – 28 by district impact, district, and source (IRM-2) Table 3.3: Share of people who have received food aid – by district 30 (IRM-1/IRM-2 comparison) Table 3.4: Share of people who have received shelter items – by district impact 31 and district (IRM-1/IRM-2 comparison) Table 3.5: Share of shelter items in top two current needs – by district impact 31 and district (IRM-2) Table 3.6: Damage assessment results and self-reported damage – by district impact 38 and district (IRM-2) Table 3.7: Satisfaction with damage assessment – by whether or not received 40 beneficiary cards (IRM-2) Table 3.8: Satisfaction with damage assessment – by how house was classified 40 in damage assessment (IRM-2) Table 3.9: Satisfaction with aid providers, among those who received aid only – 43 by district impact and district (IRM-2) Table 3.10: Types of aid received – by income band (IRM-2) 44 Table 3.11: Share of people who have received aid of different types – by caste groups (IRM-2) 45 Table 3.12: Aid received – by level of housing damage (IRM-2) 46

Table 4.1: Average amount borrowed (NPR) per borrower – by district impact 49 (IRM-1/IRM-2 comparison) Table 4.2: Average amount borrowed (NPR) per borrower – by district impact 49 and district (IRM-2) Table 4.3: Intention to borrow in the next three months – by district impact and 50 district (IRM-2) Table 4.4: Proportion borrowing and amount borrowed – by caste (IRM-2) 52 Table 4.5: Sources of borrowing among those who borrowed – by income band (IRM-2) 53 Table 4.6: Remittances as a share of main income source, impact on remittances, and 54 recovery of remittances – by district impact and district (IRM-2)

xiii Aid and Recovery in Post-Earthquake Nepal

Table 5.1: Satisfaction with local political parties – individual panel data 58 (IRM-1/IRM-2 comparison) Table 5.2: Satisfaction with local political parties – by district (IRM-2) 59 Table 5.3: Satisfaction with the central government – individual panel data 59 (IRM-1/IRM-2 comparison) Table 5.4: Current political preferences – by past votes (IRM-2) 60

Table 6.1: How safe and secure do you feel now in your community? – by current 62 shelter (IRM-2) Table 6.2: Fair distribution by VDC/municipalities – by district impact and district 66 (IRM-2)

xiv Aid and Recovery in Post-Earthquake Nepal

LIST OF CASE STUDIES

Case Study 1: Health in temporary shelters 15 Case Study 2: Young entrepreneurs in Syangja start a concrete block factory 21 Case Study 3: Trekking loses, masonry gains 22 Case Study 4: Indecision due to unclear and delayed government policy 35 Case Study 5: Rumors 39 Case Study 6: Tensions over relocation in Sinhulpalchowk 64

LIST OF MAPS

Map 1.1: Locations of surveyed districts 3 Map 1.2: Locations of qualitative research 4

xv Paljor Aid and Recovery in Post-Earthquake Nepal

Chapter 1. Introduction

Photo: Ashray Pande

1.1 Background

The lives of eight million people, almost one-third qualitative fieldwork at regular intervals in the same of the population of Nepal, were impacted by the locations. The second wave of research (IRM-2) was earthquake of 25 April 2015 and the subsequent after- conducted in 11 quake-impacted districts from Feb- shocks.1 The Post-Disaster Needs Assessment (PDNA), ruary-March 2016. The report synthesizes findings prepared by the Government of Nepal in August 2015, from both components of research. More data and found that over 8,790 people were killed and 22,300 detailed findings can be found in the qualitative and injured across 31 affected districts, 14 of which were quantitative reports, which are published in parallel.3 declared severely or crisis hit.2 Damages and losses totaled over NPR 590 billion. One year on, how are IRM is premised on the belief that the impacts of the earthquake-affected recovering? How effective major disasters such as the Nepal earthquake do not has the aid response been? And what needs remain? only manifest immediately but play out in complex and multidimensional ways over the longer run. Many This report provides findings from the second round of the direct impacts—deaths and injuries, decimated of the Independent Impacts and Recovery Monitoring houses and public infrastructure—are immediately for Accountability in Post-Earthquake Nepal (IRM) apparent. The PDNA quantified such damages and research project. IRM is a longitudinal mixed methods losses and the first round of IRM also provided infor- study, involving rounds of household surveying and mation on the level of destruction.4 Yet disasters affect

1 Government of Nepal, National Planning Commission. Nepal and Interdisciplinary Analysts (2016). Aid and Recovery in Post- Earthquake 2015: Post-Disaster Needs Assessment (Volume A: Key Earthquake Nepal: Independent Impacts and Recovery Monitoring Findings). Kathmandu 2015. (available at: http://www.npc.gov.np/ Nepal Phase 2 – Quantitative Survey (February and March 2016). images/download/PDNA_Volume_A.pdf) Throughout, the report The Asia Foundation and Democracy Resource Center Nepal (2016). notes the impacts of ‘the earthquake’, but this also includes the Aid and Recovery in Post-Earthquake Nepal: Independent Impacts impacts of the subsequent aftershocks, which affected many people. and Recovery Monitoring Nepal Phase 2 – Qualitative Field 2 Ibid. Monitoring (February and March 2016). 4 3 The qualitative field research was conducted by Democracy The Asia Foundation (2015). Aid and Recovery in Post-Earth- Resource Center Nepal (DRCN) from 20 February to 8 March quake Nepal: Independent Impacts and Recovery Monitoring 2016. The quantitative survey was conducted by Interdisciplinary Nepal Phase 1 – Quantitative Survey (June 2015). Kathmandu and Analysts (IDA) from 12 February to 11 March. The Asia Foundation Bangkok: The Asia Foundation.

1 Introduction

populations in many other ways that often take time to the relief effort has shifted from emergency assistance appear. Social, economic, and political structures and to recovery and reconstruction. institutions can be affected by such massive shocks. Understanding these deeper impacts requires tracking The pace of recovery, and the experiences of different levels of recovery, and things that are supporting or population groups, will be determined by the level hindering it, over time. of earthquake impacts, the aid response, the coping strategies employed by affected households and This second round of research builds on initial data communities, and the political and economic context collected during a first wave of research (IRM-1), in which the recovery is taking place. IRM focuses which was concluded eight weeks after the earthquake. on each of these issues at the local level to assess the 21 This report refers back to IRM-1’s data on the delivery extent to which recovery is taking place, how this of humanitarian assistance and the earliest phases of varies between groups and areas, and the causes recovery to better understand how the situation has of differences in the degree and nature of recovery evolved as households have sought to cope with the (Figure 1.1). challenges of the monsoon and winter seasons and as

Figure 1.1: Analytic framework

V I

I V

2 Aid and Recovery in Post-Earthquake Nepal

1.2 Methodology

The IRM-2 quantitative survey involved face-to-face a minimum of 350 people interviewed per district. interviews with 4,850 respondents, plus surveys Another 1,000 households across four districts were with ward leaders, across 305 wards. These were added to allow for analysis of food insecurity. The conducted in 11 districts (Map 1.1), all of which were household sample was distributed equally among men covered in the IRM-1 survey. Respondents in IRM-1 and women aged over 18. Data collection took place in: were selected using stratified randomized sampling. Nuwakot, Sindhupalchowk, Ramechhap, Gorkha, and To assess changes since IRM-1, we sought to re- Dhading (severely hit); Bhaktapur, Okhaldhunga, and interview those people who were interviewed in the Kathmandu (crisis hit); Solukhumbu and Lamjung first wave of research. This was not always possible. (hit with heavy losses); and Syangja (hit). Severely hit In total, 1,558 people were interviewed in both IRM- districts are the most affected districts, followed by 1 and IRM-2. To boost the precision of district level crisis hit districts, then hit with heavy losses districts, estimates, an additional 67 wards were selected with and then hit districts.5

Map 1.1: Locations of surveyed districts

The qualitative research involved teams conducting two factors that were predicted to affect the nature interviews, focus group discussions, and participant and speed of recovery: the degree of impact of the observation in six districts spread across different earthquake; and the degree of remoteness. earthquake impact categories: Sindhupalchok, Gorkha, and Ramechhap (severely hit); Okhaldhunga (crisis hit), Solukhumbu (hit with heavy losses), and Syangja (hit) (Map 1.2). Research teams visited 16 village development committees (VDCs) and two municipalities, with two wards studied in each. Research took three-four days per VDC and was supplemented by interviews in district capitals. 5 Affected districts were categorized based on the Nepali govern- Sampling of locations was done at three levels— ment’s PDNA. More information on the methodology is provided district, VDC, and ward—to maximize variation in in the IRM-2 survey report.

3 Introduction

Map 1.2: Locations of qualitative research

The methodology for both components of the research dence interval. Where we break down the surveyed was developed to ensure to the greatest degree possible population by impact, demographic, or other variables that findings accurately reflect conditions and views (for example, comparing the opinions of men or wom- in earthquake-affected areas. A few relevant details en, or patterns of recovery in districts with different regarding the methodology and its limitations should levels of earthquake impact) the level of accuracy of be noted. survey findings reduces. For example, the margin of error for district-disaggregated analyses is +/- 5.2%. Timing of research Additional households were added in four districts to allow for a focus on food insecurity. The results IRM-2 fieldwork was conducted from February can be generalized at an error margin of +/- 4.0% in to March 2016. During this period, large-scale these districts.6 It should be noted that these margins government reconstruction policies and schemes had of error are very small compared to most surveys, in not yet been rolled out. As such, the report does not Nepal and beyond. evaluate policies or aid provided after March 2016. Future rounds of IRM will capture more information Tracking the same areas over time on those developments. IRM is set up to see how conditions evolve over time. Confidence in findings As a result, research is conducted in the same places for each wave. However, there were some changes in The quantitative survey is representative of the full the districts studied between IRM-1 and IRM-2. IRM-1 population of the 11 earthquake-affected districts was conducted before the PDNA classification of earth- where the survey was conducted. Across the whole quake districts was completed. As such, adjustments sample, the error margin is +/- 1.4% at a 95% confi- were made for IRM-2 to make sure that research

6 Data from the Nepal Food Security Monitoring System (NeKSAP) security. Across food security categories in the four districts as a was used to identify areas with different levels of food insecurity. The whole, the error margin is +/- 4.0%; across food security categories sample was boosted in four districts (Sindhulpalchowk, Ramechhap, within each district, the error margin is +/- 7.0%. The NeKSAP data Gorkha, and Okhladunga) to allow for deeper analysis of food came from meetings held 15-30 November 2015.

4 Aid and Recovery in Post-Earthquake Nepal

Photo: Binu Sharma

aligned with the government’s classification. For the Perceptions and accurate reporting household survey, three of the 14 districts surveyed in IRM-1 (Manang, Khotang, and Dang) were dropped The information provided throughout the report is from the sample as these districts were not included in based on the reports of those interviewed. People may the PDNA as impacted districts. The IRM-1 data was have incentives to over- or under-report the level of reweighted to reflect the change in district sampling impact they experienced, and their perceptions or feel- and comparisons in this report use the reweighted ings might not accurately reflect facts in some cases. data. For the qualitative research, Ramechhap and The data and findings should be read with this in mind. Solukhumbu replaced Dolakha and Makawanpur for But the use of both qualitative and quantitative re- the second round. Going forward, no further changes search has allowed for triangulation of findings, which to district sampling are expected. strengthens our confidence that they reflect reality.

5 Introduction

1.3 Contextual changes since June 2015

A number of key contextual changes since IRM-1 was respondents reported that, while it was ongoing, the conducted have shaped recovery. fuel crisis had led to a reduction, or even stopping, of aid flows to their community (Figure 1.4). In the weeks after the disaster, as the monsoon rains began to descend in June 2015, individual citizens, the Government of Nepal, civil society, and the Figure 1.2: Effect of the crisis on access to fuel for international community overcame extraordinary cooking and transportation (IRM-2) challenges to distribute relief and temporary shelter for thousands of displaced and homeless persons 4% residing in earthquake-affected regions. 10%

However, the shift from relief to recovery and reconstruction has been slow, in large part because of delays from the government in setting up a regulatory and administrative structure to oversee the earthquake effort. It took until December 2015 for the National Reconstruction Agency (NRA) to be established. Even 35% 49% after it was inaugurated in January 2016, political wrangling over who would lead it meant that it did not become active in practice for a number of months.

The delay in creating the NRA in turn led to delays 2% in reconstruction efforts. These, combined with challenges communicating effectively with local earthquake-affected areas and persons, and multiple (sometimes conflicting) damage assessments, have Both affeted meant that policies and plans have been unclear – for ooking affeted, transport unaffeted local officials and earthquake-affected people alike. ooking affeted, transport not required As of the time of the IRM-2 fieldwork, government Transport affeted, ooking unaffeted support for the reconstruction of private houses had Both unaffeted yet to begin. In recent months the NRA has started to sign agreements with people who are eligible to receive reconstruction assistance. It is within this context that the report assesses These delays were exacerbated by political develop- how people have recovered, how the aid effort has ments in Nepal. The promulgation of a new constitu- helped, and what people have done to cope with these tion on 20 September 2015, after years of delay and po- immense challenges. litical deadlock, was a major development. But the new constitution was not welcomed by all. Dissatisfaction with elements, including the issue of federal boundary demarcation, led to protests, violent incidents, and an economic blockade along the Nepal-India border. This blockade of multiple major border crossings started in September 2015 and did not end until February 2016. This resulted in a severe shortage of fuel in the country and the scarcity of some other goods.

The shortage of fuel hampered recovery. The IRM-2 survey found that access to fuel for cooking affected two-thirds of people in earthquake-affected areas (Figure 1.2). Prices of goods increased markedly (Figure 1.3). There was also an impact on the distribution of aid. Fifty-four percent of survey

6 Aid and Recovery in Post-Earthquake Nepal

figure 5.10

Figure 1.3: Share saying the fuel crisis affected the prices paid for food types (IRM-2)

100 98% 98% 95% 93%

78% 5 70% 65%

0 ie Wheat entils ooking oil egetables Sugar et eat

Figure 1.4: Share saying protests over the new Constitution and the fuel crisis affected aid assistance in people’s ward – by district impact (IRM-2)

41% 38%

25 25%

19% 19% 16% 15%

1 11%

7% 8%

1% 0 0 es, aid es, less aid es, more aid o ever reeived ont know ompletel stopped than before than before an aid assistane

Severel hit All distrits

7 Introduction

1.4 Report structure

This report covers a number of areas:

• The impacts of the earthquake and critical • Coping strategies. Chapter 4 looks at how house- needs of affected households and communities. holds have tried to cope with earthquake impacts Chapter 2 considers the critical needs of the through financial behavior, migration, and other earthquake-affected, along with the scale of means; damage to property, public infrastructure and facilities, and disruption of livelihoods; • Politics and leadership. Chapter 5 reviews the extent to which the earthquake and aid response • Earthquake aid. Chapter 3 details the nature of have affected political party activities, roles aid provided and how this has changed over time, and levels of influence, the emergence of new people’s experiences and levels of satisfaction leadership, and political preferences; with assistance received and with those providing it, and the coordination and transparency of aid • Social relations and conflict. Chapter 6 details the distribution; impacts of the disaster and response on security, sources of conflict, and social cohesion.

Analysis of the differing impacts on different popu- The report concludes with a summary of findings and a lation groups, differing patterns of recovery, and the discussion of implications for aid and recovery efforts extent to which groups are vulnerable, is provided moving forward. throughout.

8 Aid and Recovery in Post-Earthquake Nepal

Chapter 2. Earthquake Impacts and Needs

Photo: Alok Pokharel

IRM-1 outlined the immense physical impacts of the Almost one year on from the earthquake, many of earthquake, with most housing in severely hit districts the affected are still living in temporary shelter and destroyed or badly damaged. In the emergency period, urgently need support to help them move into more there was a vast array of needs for affected people, robust housing. Livelihoods are recovering but support including for temporary shelter, food, and cash. to farmers is needed.

2.1 Key needs

One year on from the earthquake, the greatest the 2015 monsoon season (IRM-1 three-month needs); needs in affected areas are cash, shelter, and and needs in February-March 2016 (IRM-2 immediate food. needs). The data show that cash is becoming an ever more important priority for people (41% in IRM-1, In IRM-1, conducted just after the earthquake, the 48% as an expected need after IRM-1, and 69% in most prominent immediate need reported by survey IRM-2). Qualitative data also suggest that cash and respondents was for cash, with 41% citing it as credit are increasingly top priorities for earthquake- amongst their top two priority needs and 48% saying affected people. it was a top need over the next three months. The next most reported immediate need was for food (30%), Over one-third of respondents identify reconstruction followed by clean drinking water (5%). Prioritization materials as a top-two priority in IRM-2. A much of shelter was low and spread across the various larger share of people say CGI is a priority immediate shelter options fairly evenly: tents (6%), tarps (7%), need than in the past (16% in IRM-2 compared to corrugated galvanized iron sheets (CGI) (5%), and 5% in IRM-1), although this is less than the 26% of “houses” (2%). IRM-1 respondents who prioritized CGI as a three- month need. This reflects the fact that tarps have One year on from the earthquake, cash and shelter been distributed in large shares to households (71% in have become higher priority needs for affected people. IRM-1 and 47% in IRM-2, see Chapter 3), while CGI Food, while still important, is prioritized less than distribution has been lower (7% in IRM-1 and 21% in in IRM-1. Figure 2.1 compares needs prioritized by IRM-2). Food, while still an important priority, has respondents across three time periods: needs in June declined in importance since IRM-1. The importance 2015 (IRM-1 immediate needs); expected needs during of water, both for drinking and for other household

9 Earthquake Impacts and Needs

Figure 2.1: Priority needs (share in top two needs) – all districts figure 51(comparison IRM-1 immediate needs, IRM-1 three-month needs, IRM-2 immediate needs)

69% ash 48% 41% eonstrution 34% 0% materials 0% 18% Food 17% 30% 16% I 26% 5% 7% rinking water 5% 5% 5% ouse 2% 2% 4% Water for household 3% 2% 3% Farm implements 5% 2% 3% ivestok 0% 0% 2% Tarps 4% 7% 2% Sanitation 2% 1% 2% edial 6% 2% 1% Tent 2% 6% 0 1 5 5 0

I-2 immediate I-1 months I-2 immediate

* Data on reconstruction materials and livestock as needs were not collected in IRM-1. Blankets, clothes, and fuel are not shown in the figure, because shares prioritizing these (for both IRM-1 and IRM-2) are negligible. uses, has increased since IRM-1. Other needs, such Eighty-five percent of people in the most affected as livelihoods inputs and medicine, are much less districts say they need cash and over half in crisis hit commonly cited. and hit with heavy losses districts also say they need financial assistance Table( 2.1). Almost half in severely These findings were confirmed by the qualitative hit districts say they need reconstruction materials, research which found that support for housing with almost one-quarter of those living elsewhere reconstruction was most commonly cited, with support stating the same. for improving temporary shelter also mentioned. The need for cash grants or credit was also a frequently Geological assessments to determine land- voiced need (the third most common across the areas slide risks were seen as a critical need in some studied). The second most frequent priority in the locations. qualitative study was support for water and sanitation infrastructure, a need identified by 11% of respondents The qualitative research found communities who (7% drinking water and an additional 4% water for the rated the need for geological assessments as a very household) in the quantitative research. Respondents high priority because of perceived landslide risks.8 who prioritized water were concentrated in severely hit These communities emphasized the need to determine districts, where 17% said it was a top priority.7 whether land was safe to live or work on. While it is unclear what the actual level of risk is, in all VDCs The need for cash and building materials is where research was conducted assessments have yet greatest in severely hit districts but also high to materialize. elsewhere.

7 Food featured less prominently in the qualitative research, most 8 Locations included Syaule VDC (ward 8) in Sindhupalchowk and likely due to differences in sampling methodology. Barpak VDC (wards 2 and 5) in Gorkha.

10 Aid and Recovery in Post-Earthquake Nepal

Photo: Alok Pokharel

Table 2.1: Priority immediate needs (share in top two needs) – by district impact (IRM-2)

Hit with heavy Severely hit Crisis hit Hit Total losses Cash 85% 53% 65% 40% 69% Reconstruction materials 47% 24% 22% 23% 34% Food 23% 13% 17% 7% 18% CGI 18% 9% 22% 9% 16% Drinking water 11% 5% 2% 4% 7% A house 11% 1% 0% 0% 5% Water for household 6% 2% 1% 1% 4% Farm Implements 5% 2% 2% 1% 3% Livestock 4% 2% 3% 1% 3% Blankets 2% 1% 6% 2% 2% Tarps 3% 1% 2% 3% 2% Medical 2% 2% 1% 3% 2% Sanitation 3% 1% 0% 1% 2% Clothes 1% 2% 1% 1% 1% Tent 1% 1% 2% 1% 1% Fuel 0% 2% 0% 0% 0%

VDC and district government officials appear orities, although they also highlighted livelihood sup- to have a good understanding of local needs. port as a means to aid earthquake victims’ recovery.

In general, the needs cited by local officials matched The number of people stating they do not need those expressed by citizens, with support for building aid any more has fallen since IRM-1. reconstruction (including guidelines), cash and pro- visions for soft loans, and rebuilding of community Two months after the earthquake, 24% of people said infrastructure commonly mentioned during the qual- they did not need aid. In IRM-2, the figure is 21%. itative fieldwork. However, food was not mentioned This suggests that some people have realized that the as a priority by government officials despite featuring challenges of recovering have been greater than they prominently in the household survey results. Non-gov- initially expected. In severely hit districts, just 1% of ernmental organizations largely identified similar pri- people say they do not need aid.

11 Earthquake Impacts and Needs

2.2 Shelter

As of March 2016, almost 80% of people in people have been able to stay on their own land, 6% severely hit districts were living in temporary in severely hit districts have had to build temporary shelters. shelters on other people’s land. This displacement is highest in Sinhupalchowk, where 16% of respondents The emphasis on shelter as a critical need is clearly were living in temporary shelter constructed on other’s linked to the large number of households who are still land (Table 2.2). Less than 1% of survey respondents living in temporary shelters (Figure 2.2). While most were living in temporary shelter on public land.

figure 5

Figure 2.2: Where people are living now – by district impact (IRM-2)

100 95% 91%

80% 81% 5

25 19% 16%

8% 4% 1% 2% 1% 1% 1% 0 0% 0% 0% Severel hit risis hit it with heav losses it

wn house Self-onstruted shelter eighborfriend ented house

Table 2.2: Where people are living now – by district (IRM-2) Own house house Neighbor’s house Friend’s Self-constructed shelter on own land Self-constructed shelter on other land people’s Self-constructed shelter on public land Community shelter Rented accommodation

Severely hit 19% 1% 0% 74% 6% 0% 0% 0% Dhading 20% 1% 0% 77% 2% 0% 0% 0% Gorkha 45% 2% 0% 50% 3% 0% 0% 0% Nuwakot 9% 0% 0% 84% 5% 2% 0% 0% Ramechhap 16% 0% 0% 80% 4% 0% 0% 0% Sindhupalchowk 6% 0% 0% 78% 16% 0% 0% 0% Crisis hit 81% 2% 0% 14% 2% 0% 0% 1% Bhaktapur 79% 2% 0% 13% 2% 1% 0% 3% Kathmandu 90% 1% 0% 8% 0% 1% 0% 0% Okhaldhunga 76% 2% 0% 21% 1% 0% 0% 0% Hit with heavy losses 91% 0% 0% 8% 1% 0% 0% 0%

12 Aid and Recovery in Post-Earthquake Nepal Own house house Neighbor’s house Friend’s Self-constructed shelter on own land Self-constructed shelter on other land people’s Self-constructed shelter on public land Community shelter Rented accommodation

Lamjung 92% 0% 0% 8% 0% 0% 0% 0% Solukhumbu 91% 1% 0% 7% 1% 0% 0% 0% Hit 95% 1% 0% 3% 1% 0% 0% 0% Syangja 95% 1% 0% 3% 1% 0% 0% 0% All districts 56% 1% 0% 39% 3% 0% 0% 1%

Most temporary shelters are made from either CGI shelter aid in both IRM-1 and IRM-2, yet continues alone (40%) or CGI with wood or bamboo (47%). Four to have one of the highest levels of immediate percent are in shelters made entirely from bamboo, 3% need for shelter aid. This reflects the fact that most in wooden shelters, and 2% in cow sheds.9 shelter aid so far has been for temporary shelter (Chapter 3). While 29% of people in the district Immediate needs for shelter are greatest in have received reconstruction materials, 34% say Nuwakot, Ramechhap, and Sindhupalchowk. that such materials are an immediate priority need and another 39% say that a house is a top priority. These three districts were highly affected by the Demand for reconstruction materials is particularly earthquake, with over 90% of houses damaged high in Nuwakot and Ramechhap, both of which have or destroyed in each. Two-thirds of people report received almost no materials so far. Demand is also shelter as an immediate need in these districts. high in Okhaldhunga. Sindhupalchowk has received the highest level of

Table 2.3: Share of people reporting shelter as priority immediate need – by district impact and district (IRM-2)

Shelter as priority immediate need Housing District Reconstruction Shelter damage (%) Tent Tarps CGI House materials (total) Severely hit 4% 1% 18% 47% 11% 72% 94% Dhading 1% 0% 20% 39% 6% 61% 97% Gorkha 2% 0% 15% 35% 11% 58% 89% Nuwakot 1% 0% 21% 69% 0% 84% 97% Ramechhap 14% 2% 18% 59% 1% 79% 90% Sindhupalchowk 1% 1% 19% 34% 39% 79% 97% Crisis hit 1% 1% 9% 24% 1% 33% 49% Bhaktapur 1% 1% 5% 16% 3% 24% 60% Kathmandu 0% 0% 3% 17% 0% 19% 26% Okhaldhunga 3% 2% 20% 40% 0% 57% 60% Hit with heavy losses 3% 0% 22% 22% 0% 43% 44% Lamjung 1% 0% 11% 23% 0% 33% 35% Solukhumbu 6% 0% 33% 21% 0% 54% 52% Hit 3% 2% 9% 23% 0% 32% 21% Syangja 3% 2% 9% 23% 0% 32% 21% All districts 3% 1% 16% 34% 5% 53% 66%

9 See Figure 5.4 in IRM-2 survey report.

13 Earthquake Impacts and Needs

Most people were able to make repairs to their they made repairs but these were not sufficient for the shelters to get them through the winter. winter. Five percent were unable to make repairs at all (Table 2.4). People in temporary shelters in less af- Of those still living in temporary shelters, 72% said fected areas were less likely to have been able to make they were able to make repairs to their shelter that sufficient repairs for winter. This is most likely due to made it sufficient for the winter. Another 21% said the relatively low levels of aid received in these areas.

Table 2.4: Were you able to make sufficient repairs to your shelter for the winter? – by district impact and district, those living in temporary shelters (IRM-2)

Was able Was able to Was able to Was not Housing to repair make repair complete- able to was not and made it but not Refused ly fix the repair the damaged sufficient for sufficient for house house at all winter winter Severely hit 0% 1% 75% 21% 2% 0% Dhading 0% 1% 98% 1% 0% 0% Gorkha 0% 1% 86% 12% 2% 0% Nuwakot 0% 0% 60% 39% 0% 0% Ramechhap 0% 1% 77% 21% 1% 0% Sindhupalchowk 0% 2% 64% 27% 7% 0% Crisis hit 1% 5% 50% 22% 21% 1% Bhaktapur 0% 2% 45% 27% 25% 0% Kathmandu 6% 16% 50% 16% 9% 3% Okhaldhunga 0% 3% 53% 21% 24% 0% Hit with heavy losses 0% 14% 47% 23% 16% 0% Lamjung 0% 0% 59% 30% 11% 0% Solukhumbu 0% 27% 37% 17% 20% 0% Hit 0% 8% 46% 0% 46% 0% Syangja 0% 8% 46% 0% 46% 0% All districts 0% 2% 72% 21% 5% 0%

However, poor conditions in temporary people whose homes had suffered significant damage shelters have led to much hardship. lived through the monsoon and winter in temporary individual or community shelters. Some individuals With the exception of Syangja district, in all wards chose to move back into their damaged or only visited in the qualitative research the majority of partially repaired homes. It was reported that those

Photo: Amanda Gurung

14 Aid and Recovery in Post-Earthquake Nepal

who returned to their homes, or who managed to build and other ailments. Children, the elderly, and preg- durable semi-permanent shelters of wood or stone, nant women in temporary shelters were particularly did not suffer as much illness or physical hardship as vulnerable to sickness during both the monsoon and those in lower quality temporary shelter. winter. Ill health among the earthquake-affected was compounded by the fact that across the 36 wards Illness and extreme physical discomforts were com- visited, 25 health posts had been either partially or mon hardships endured by those unable to make their fully damaged by the earthquake. Women faced issues temporary shelter ready for monsoon and winter. relating to insecurity and discomfort in temporary Common challenges included leaking tarpaulins, shelters, exposure to danger as part of household drafty shelters, damp clothes, leech bites, and lower duties, and increased burdens taking care of children quality food, swollen hands and faces, fevers, coughs, and the elderly in challenging conditions.

Case Study 1: Health in temporary shelters A widow in Baruneshwor Ward 1 has lived also been falling sick repeatedly. She said, “if we alone since her husband died 13 years ago. had a proper house, we would not be wasting so While other families in the village started much money on medicines.” constructing makeshift bamboo huts after the quake, she had more difficulty securing shelter. The senior health assistant of Rampur health She borrowed money to supplement the NPR post told researchers that the number of 15,000 given by the government and built a patients has more than doubled following shelter made of bamboo, tarp, and CGI sheets. the quake. “Most patients had skin diseases She has been frequently ill during the eight and they also had psychological problems like months she has lived there. Common visits to fear and anxiety. After January, we treated the hospital in the district headquarters has many patients with pneumonia, coughs, fever, meant she has not been able to work regularly. and problems related to breathing.” He also She is thus losing out on daily wages while also noted an increase in malnutrition cases. “I spending an unexpected amount of money at treated 19 children with malnutrition after the hospital. Due to her health needs, she spent the quake. Since parents might be unable to the NPR 10,000 winterization cash given by work regularly, their regular income has been the government on medical treatment. “I suffer disrupted. This might have resulted in the lack from asthma,” she said, “which has worsened of nutritious food at home.” This health worker after I started living in this cold hut.” expressed further fears that health risks could escalate once summer began as unclean shelters Quake victims in Katunje are still in makeshift became breeding grounds for mosquitoes, shelters and many children and the elderly which transmit many diseases. have fallen ill. One old women told researchers: “my 22-month-old granddaughter has been The chief of Katunje health post pointed to ill constantly since we started living in the the lack of proper living spaces as the cause shelter. Even my adult daughter caught a skin of an increase in health-related issues among allergy after we shifted to the hut. Others have the people in the VDC: “many children suffer fevers.” A 55-year-old man from the same VDC pneumonia because they do not have enough recounted the troubles he had suffered since warm clothes, proper food, and warm places to his wife fell ill in their temporary hut. He took sleep. Children from more than half of the Dalit his wife to Biratnagar where, he believes, the families of Wards 1 and 8 have fallen ill.” He also health facilities are better and the hospital is pointed out that issues of malnutrition in Dalit better equipped. “The doctor asked us to avoid children were not common before the quake, cold, smoke, and dust. How is that possible in but now, “two children in Ward 8, one in Ward the hut? The floor is of mud, the wall and the 2, and one in Ward 5 have been badly affected ceiling are made of CGI. How can it be warm? by severe malnutrition. Before the quake, there We use firewood for cooking which means we were only four cases of malnutrition in the cannot avoid smoke.” A mother in Katunje entire VDC and now we have traced 17 cases of Ward 1 complained about her infant daughter’s malnutrition. The majority of these children are frequent fevers. The adults in the family have from the Dalit families.”

15 Earthquake Impacts and Needs

2.4 Food

While food has declined in importance since food needs among the non-severely hit districts are IRM-1, large amounts of people still say it is Okhaldhunga, where food aid has fallen to very low one of their top two priority immediate needs. levels, and Solukhumbu, which is the poorest of the surveyed districts (Figure 2.3). The latter district has Current demand for food is the highest in the severely the highest share reporting food as an immediate hit districts (23%) with 13% in crisis hit districts, 17% need, followed by Gorkha and Sindhupalchowk, the in hit with heavy losses districts, and 7% in the hit worst affected district in terms of damage to homes district reporting it as an immediate priority need. and livelihoods. Current and future food needs are the The only districts with high levels of immediate lowest in Lamjung, Kathmandu, and Syangja.

figure 55 Figure 2.3: Share of people reporting food as among their top two needs immediately and in three months – by district (IRM-2)

40 39%

32% 31% 29% 0 28%

24% 23% 22% 21% 19% 20 18% 18%

14% 12% 10% 9% 10 8% 7% 7% 6% 3% 2% 0 hading orkha uwakot amehhap SindhupalhowkBhaktapur athmandu khaldhunga amung Solukhumbu Sanga

Immediate Three months

Staples (rice and wheat) are the greatest food (Table 2.5). These patterns were roughly the same needs followed by condiments. when households were asked what kinds of food aid they would need in three months. The demand for all types of food is the highest in the severely hit districts followed by the crisis hit districts

Table 2.5: Share of people reporting food as priority immediate need – by district impact (IRM-2) Rice/wheat Readymade food salt, Sugar, etc. Lentils Vegetables Meat Food (total)

Severely hit 15% 2% 7% 0% 1% 1% 23% Crisis hit 11% 1% 4% 1% 1% 0% 13% Hit with heavy losses 10% 5% 6% 0% 0% 0% 17% Hit 5% 1% 0% 0% 0% 0% 7% All districts 12% 2% 5% 0% 1% 0% 18%

16 Aid and Recovery in Post-Earthquake Nepal

The number of people receiving food aid has quake impacts. The share of people reporting that food fallen across almost all districts and this has consumption has decreased a lot is fairly low (7%) and led to decreases in food consumption. most people have been able to cope with decreasing food aid. Dhading, however, which had the sharpest Wherever food aid has fallen, a larger share of peo- decline in households receiving food aid (93% to 38%), ple report that consumption has slightly decreased has the highest number of households reporting sig- since the 2015 monsoon compared to other districts nificantly declining food consumption since the 2015 (Table 2.6). This suggests that food aid is affecting monsoon (15%). This correlation is also seen in Sind- overall consumption levels, despite households also hupalchowk (24% reported that consumption declined reported borrowing for food as they coped with earth- slightly or a lot), Gorkha (7%), and Ramechhap (6%).

Table 2.6: Food aid, borrowing, food consumption, and current need for food – by district impact and district (IRM-1 and IRM-2)

Change in food consumption Food aid since monsoon (June 2015) (IRM-2) Borrowing Immediate for food need now (IRM-2) (IRM-2) IRM-1 IRM-2 Difference (IRM-1 – IRM-2) Increased a lot Increased slightly No change Decreased a lot Decreased slightly

Severely hit 93% 65% 28% 35% 3% 26% 60% 9% 2% 23% Dhading 93% 38% 55% 44% 1% 10% 73% 15% 1% 18% Gorkha 89% 50% 38% 18% 6% 17% 70% 7% 0% 29% Nuwakot 96% 87% 10% 33% 2% 44% 53% 1% 0% 22% Ramechhap 89% 67% 22% 49% 7% 47% 41% 6% 0% 19% Sindhupalchowk 100% 83% 17% 26% 1% 14% 62% 14% 8% 28% Crisis hit 26% 7% 18% 40% 9% 26% 61% 5% 0% 13% Bhaktapur 34% 11% 23% 18% 7% 24% 64% 5% 0% 12% Kathmandu 9% 9% 0% 15% 11% 17% 68% 3% 0% 7% Okhaldhunga 34% 2% 32% 54% 8% 35% 50% 7% 0% 21% Hit with heavy losses 8% 24% -16% 21% 3% 25% 69% 3% 0% 17% Lamjung 6% 15% -9% 27% 3% 22% 69% 4% 0% 2% Solukhumbu 10% 33% -23% 16% 3% 28% 68% 1% 0% 31% Hit 3% 4% -2% 37% 1% 39% 58% 3% 0% 7% Syangja 3% 4% -2% 37% 1% 39% 58% 3% 0% 7% All districts 53% 37% 17% 35% 5% 27% 60% 7% 1% 18%

Food aid appears to reduce the need to borrow Food aid has been effectively targeted at areas for food. with higher food insecurity.

Table 2.6 also demonstrates a negative relationship Using levels of food insecurity reported by the Nepal between level of food aid and level of borrowing (-23%) Food Security Monitoring System (NeKSAP), and the in the five poorest five districts (Dhading, Nuwakot, additional sample in four districts (Sindhupalchowk, Ramechhap, Okhaldhunga, and Solukhumbu). This Ramechhap, Okhaldhunga, and Gorkha), IRM-2 was suggests that food aid reduces the need to borrow, able to carry out an analysis of how areas with different and could possibly imply that a continuation of the levels of food insecurity were experiencing food aid trend of reducing food aid could result in higher rates since the earthquake. More food insecure areas have of borrowing for food in the future. received higher levels of food aid (Table 2.7). Severely food insecure areas have had near full coverage of food There is also a very high negative correlation between aid (96%). Volumes of food aid have also been larger the proportion of borrowers who are taking loans for at successively higher levels of food insecurity and the food and the shares within the district who cite food highest volumes are in the severely insecure category as a top current need in the poorest districts (- 86%). (103 days of stock for the household), double that in This suggests that where there has been less borrowing the highly insecure category (52 days). for food as a means to cope, there is also a greater immediate need for food.

17 Earthquake Impacts and Needs

Table 2.7: Food aid, borrowing, consumption, and need for food – by level of food insecurity (IRM-2)

Food aid Change in food consumption Food as need NeKSAP food security classification receiving Share food aid Quantity of food aid (days of food per household) a lot Increased Increased slightly Same as before a Decreased lot Decreased slightly for food Borrowed Immediate Three-month

Minimally food insecure 28% 22 11% 23% 60% 6% 0% 18% 21% 21% Moderately food insecure 55% 41 4% 27% 56% 9% 3% 19% 19% 24% Highly food insecure 62% 52 4% 36% 52% 8% 1% 28% 22% 28% Severely food insecure 96% 103 0% 17% 61% 15% 8% 14% 33% 42% Total 55% 52 5% 28% 56% 8% 2% 21% 22% 26%

However, levels of food aid to severely food that while food aid has been targeted effectively insecure areas are not enough to prevent based on food insecurity, volumes of food aid are not reductions in food consumption. enough. This is also reflected in the fact that food as a share of the top two needs is the highest in severely Despite receiving higher levels of food aid, the severely insecure regions now (33%) and in three months’ insecure category has the largest share reporting time (42%), despite these places receiving the most that food consumption has decreased a lot (15%) food aid. and slightly (8%) since the monsoon. This suggests

2.4 Livelihoods

The earthquake had major impacts on peo- people have multiple sources of income. For example, ple’s livelihoods. farmers may also undertake daily wage labor at quiet points in the agricultural cycle. The ability to shift to IRM-1, conducted in June 2015 shortly after the other forms of work to sustain income has meant that earthquake hit, found that all occupational groups wholesale livelihood change is not necessary. had suffered negative effects. Overall, 57% said that their main sources of income were affected by the Business people and daily wage laborers were earthquake, with shares affected declining with each particularly affected but farmers make up the successive impact category.10 Seventy-six percent of largest share of those affected. people were affected in the severely hit districts, 56% in the crisis hit districts, 34% in hit with heavy losses As was the case two months after the earthquake districts, and less than 10% in Syangja, the hit district (IRM-1), those who work in business or who are daily (Figure 2.4). wage labors are the most likely to say their income was negatively affected Table( 2.8). Given that many wage Despite the widespread livelihoods impacts, laborers earned relatively little, this group may be very few people have changed their occupa- particularly vulnerable. Over half of those who make tions as a response. money from renting property or from farming their own land also report that their income was negatively Only 0.6% of the total population, and 1% or less in impacted. Those who work for the government and, each district, have changed their livelihoods because especially, those for whom remittances are a primary of the earthquake. In large part this is because most source of income, were less likely to see their income negatively affected. These people also tend to have higher incomes than many others.

10 This includes people who said their income was either severely While a lower percentage of farmers were impacted or somewhat affected. than business owners or wage laborers, 77% of survey

18 Aid and Recovery in Post-Earthquake Nepal

respondents listed farming their own land as a main there were more farmers affected than any other source of income meaning that in absolute numbers income category.11 figure 21

Figure 2.4: Share of people whose source of income was affected by the earthquake – by district impact (IRM-2)

57% All distrits

9% it

34% it with heav losses

56% risis hit

76% Severel hit

0 20 40 60 0

Table 2.8: Share of people whose source of income was affected by the earthquake – by district impact and district (IRM-2) Farming own land Farming other's land Daily wages Own business Remittances Private company Government service Rent Livestock

Severely hit 75% 56% 69% 74% 12% 37% 25% 71% 43% Dhading 57% 11% 38% 72% 8% 44% 36% 100% 20% Gorkha 76% 56% 77% 68% 13% 20% 23% 33% 49% Nuwakot 85% 40% 75% 64% 9% 20% 15% - 27% Ramechhap 69% 85% 72% 79% 18% 50% 32% 50% 40% Sindhupalchowk 90% 86% 83% 87% 12% 50% 22% 100% 77% Crisis hit 41% 33% 60% 72% 12% 39% 31% 59% 20% Bhaktapur 50% 21% 65% 86% 14% 48% 47% 53% 44% Kathmandu 52% 75% 75% 86% 20% 60% 26% 66% 0% Okhaldhunga 20% 3% 40% 43% 2% 9% 20% - 16% Hit with heavy losses 33% 8% 20% 44% 5% 25% 26% 33% 52% Lamjung 7% 15% 9% 23% 3% 0% 7% 0% 4% Solukhumbu 60% 0% 31% 65% 8% 50% 45% 66% 100% Hit 9% 0% 11% 6% 6% 0% 3% - 17% Syangja 9% 0% 11% 6% 6% 0% 3% - 17% All districts 53% 43% 59% 72% 9% 41% 25% 57% 40%

11 See Table 2.1 in IRM-2 survey report for figures on the occupation of people in earthquake-affected areas.

19 Earthquake Impacts and Needs

Figure 2.5: Housing damage and livelihood Livelihoods in urban areas were more affected. figure 25damage – by urban/rural (IRM-2) The share of urban households reporting that 0 livelihoods were somewhat or totally affected by the earthquake is 66% against 55% in rural areas. This is 71% despite rural areas having suffered a far greater share 66% of housing damage (71%) than urban areas (38%) – 60 Figure 2.5. The differences in housing damage is clearly 55% linked to the use of different housing materials.12 The higher share of damage to urban livelihoods is likely a result of more people being engaged in businesses, 40 38% private company employment, and wage employment relative to rural areas.

20 Among the three major livelihoods—farming, livestock-rearing, and businesses—the latter have recovered the most in the last three months (Figure 2.6).13 0 Seventy percent of those in business whose occupation rban ural was negatively affected report that their livelihood situation has improved over the past three months ousing damage ivelihood damage compared to 57% engaged in livestock rearing and 48% who farm their own land. Even larger proportions of those who work in government or who are daily wage workers, whose income was affected, have seen improvements in the last three months. But far fewer people work in these occupations (2.3% and 2.7%, figure 26 respectively).

Figure 2.6: Share of affected sources of income that have improved in the past three months – by source of income, all districts (IRM-2)

57% ivestok 65% ent 77% overnment servie 65% rivate ompan 53% emittane 70% wn business 74% ail wages 41% Farming other land 48% Farming own land 0 20 40 60 0

Most farmers have restarted agricultural activities but face remaining challenges.

In the immediate aftermath of the earthquake, 12 Eighty-four percent in rural areas had homes with walls made farming was affected because entire communities with mud mortar against only 14% in urban areas; and 93% in shifted their focus to the construction of temporary urban areas used either galvanized or zinc roofs against only 63% shelters. Other challenges included the fear of in rural areas. aftershocks and landslides, displacement from their 13 Time period: first quarter of 2016.

20 Aid and Recovery in Post-Earthquake Nepal

original settlements, and the inability to plant crops land have struggled to return to their farms. Together, due to a lack of seeds or disruption of water sources these challenges have meant that agricultural workers in many highly impacted wards. The loss of draft face particular challenges in recovering, with the animals, particularly in Sindhupalchowk, also affected earthquake’s impacts still felt my farmers. productivity, especially when it was time for farmers to plough their fields. Displaced farmers were sometimes Businesses are recovering but those related forced to sell their draft animals immediately after the to tourism have struggled. earthquake at lower than market prices, and later had to purchase other draft animals to restart cultivation, In contrast, while fewer people work in business in generally paying a higher price. highly affected areas, many of those who do suffered substantial economic losses. Impacts on businesses The qualitative research found that most farmers were mainly related to damages to business-related restarted their agricultural activities after the 2015 infrastructure. In the highly-hit wards in Gorkha and monsoon, in some cases before. This is true even in the Sindhupalchowk districts, many shops and hotels most severely affected areas of Gorkha and Sindhupal- sustained major damage and took some time to re- chowk. Increasing numbers of farmers started working vive. For those in business, and related skilled labor, their fields again after the monsoon season had ended, the speed of recovery has been dependent on the partly because the fear of aftershocks and earthquake specific damage suffered, but the field work showed gradually subsided, and the risk of rain-triggered signs of general recovery, and the quantitative data landslides also decreased, but also because farmers in Figure 2.6 supports the finding that businesses are could not afford to abandon their work for too long. now recovering in earnest. In some cases, entrepreneurs have built new or expanded their businesses Farmers were less likely than others, however, to see to take advantage of the opportunities afforded by recovery of their livelihoods in the first three months reconstruction (Case Study 2). The clear exception to of the year. The time out from farming meant that this has been the tourism industry. At the time of the many lost a harvest. The lack of capital to fully replace field research, tourism, particularly in Solukhumbu livestock affected many. Many farming inputs, such as and Gorkha districts, was still struggling to recover seeds, were lost. For some, the fear remains that land from the losses incurred by the earthquake. they farm is unsafe. And those living away from their

Case Study 2: Young entrepreneurs in Syangja start a concrete block factory A group of young entrepreneurs in recent research led to the idea for concrete blocks, and months have started a factory to manufacture coincidently one of the entrepreneurs had a earthquake-proof concrete blocks. This group relative who ran a concrete block factory in was looking into earthquake resistant building Butwal who could provide guidance on how materials on the internet when they came to start manufacturing. This helped the young across a model for concrete blocks that required men recognize that their new venture would not relatively little capital. By their own admission, require a large up-front investment. The group it was the earthquake that encouraged them to registered a company in June 2015 and have pursue this technology. invested NPR 800,000 to date, including the purchase of a machine from Thailand for NPR One of the leaders of the group was unemployed 300,000. They took training on manufactur- before the earthquake, and had been looking to ing concrete blocks and started producing and start a business but had not yet found the right selling blocks in September 2015. The group idea. When the earthquake damaged and de- has been encouraged by the demand for the stroyed many houses in their village, including blocks, with new orders coming from surround- their own, they began thinking about ways to ing villages. They expect increased demand as rebuild in a safer way as houses made of mud, reconstruction efforts intensify. bricks, and stones were deemed unsafe. Internet

21 Earthquake Impacts and Needs

There have been positive impacts on wage engaged in construction and unskilled manual labor labor. had opportunities after the earthquake to help clear away damaged structures, build temporary shelters, Wage labor, one of the major sources of livelihood and work on some of the limited reconstruction activ- across the affected districts, was also affected in the ities that had started. As demand for labor increases, immediate aftermath of the earthquake, but impacts there is some evidence that wages for day laborers are were mixed. Skilled laborers such as tailors, black- also increasing. The qualitative research found that in smiths, and factory workers suffered economic losses some locations there has been a doubling of wages.14 due to disruption of their routines. Agriculture-related The increase in wages for construction jobs led to some day labor was affected in the months after the earth- farmers choosing to engage in more daily wage labor quake, but this is slowly normalizing now. Laborers to supplement their regular income from farming.

Case Study 3: Trekking loses, masonry gains At least one person in each household in Kerung and in the wage of the masons. A mason from VDC is either a mason or a carpenter. Another Kerung 1 said that there had been at least a very important source of livelihood is trekking 40% increase in the price of timber since the and tourism. On average, one person from each monsoon, and a 30-60% increase in the daily household in the VDC also spends between wages of masons. Some masons from Kerung two and three months each year working are even now traveling outside of the district as a trekking guide or a porter. According to build houses in areas where wages are much to a low level porter, an individual can earn higher. approximately NPR 70,000 per season. The total amount for the entire VDC is a significant Although trekking has taken a major hit due to sum. During the off-season, these people work the earthquake, it is hoped that the sector will in secondary occupations such as farming and bounce back to near normal levels in the high small businesses. tourist season in the Fall of 2016. If this return comes, the economy of the VDC should see a After the earthquake, with the increase in major boost. The availability of a high number demand for reconstruction of both semi- of masons in this locale has the potential to be permanent shelters and permanent houses, a great local asset as reconstruction efforts start there has been an increase in the price of timber in earnest.

2.5 Infrastructure and service delivery

The earthquake also impacted public services. people reported that other services also suffered Households reported that schooling suffered the similar setbacks: 35% for drinking water, 31% for most significant setbacks from the earthquake. One- motorable roads, and 28% for medical facilities. third of respondents in IRM-1 stated that access to schools worsened a lot because of the earthquake and There has been little progress on repairing or 36% stated that somewhat worsened.15 Other public rebuilding public infrastructure. services were relatively less affected but respondents still noted negative impacts. While 45% of the IRM- In Ramechhap, only 13% of damaged public infrastruc- 1 sample responded that electricity had worsened ture in the wards visited by the qualitative research (either a lot or somewhat), a significant number of team had been rebuilt, or was in the process of being

14 In Bamtibhandar VDC in Ramechhap district, wages jumped VDC in Okhaldhunga district, daily wages rose from NPR 600 to from NPR 600 to NPR 1,200 per day for skilled laborers such as NPR 1,000 for skilled laborers. masons and carpenters, and from NPR 450 to NPR 900 for unskilled 15 IRM-1 survey report, p. 17. laborers. Similarly, in Barpak VDC in Gorkha district, and Katunje

22 Aid and Recovery in Post-Earthquake Nepal

rebuilt, when the research team visited. In Gorkha Figure 2.7 shows that water and irrigation infrastructure and Solukhumbu, the figure is 15%. In Okhaldhunga, was the most likely to have been rebuilt by IRM-2, with none of the damaged public infrastructure had been, the repair of health posts receiving the least amount or was being, rebuilt. Sindhupalchowk was the ex- of attention at the time of the field research. ception, with 45% of damaged public infrastructure in the wards visited either rebuilt or in the process of being rebuilt. figure 21

Figure 2.7: Reconstruction of damaged infrastructure (IRM-2 qualitative research)

100

5 14 5 8 17 28 50

2 25 1 7 1 2 2 4 1 0 ealth post ther government Shools offie Water and offies irrigation soures

ebuilt ot rebuilt Being rebuilt

figure 512

Figure 2.8: Satisfaction with public services (IRM-2)

100 9% 9% 11% 17% 6% 21% 6% 5% 4% 5 6%

85% 85% 84% 79% 73%

0 letriit edial Water Shool oad

Satisfied eutral issatisfied

23 Earthquake Impacts and Needs

figure 51

Figure 2.9: Dissatisfaction with services – individual panel data (IRM-1/IRM-2 comparison)

11% otorable roads 9%

Shools 5% 8%

edial failities 9% 6%

Water 21% 11%

letriit 14% 11%

0 6 12 1 24

I-2 I-1

People are most dissatisfied with water and Levels of dissatisfaction have increased for electricity services. all services, except for schools.

Given the limited effort to rebuild public infrastruc- Overall, 1,558 individuals across the 11 districts were ture to date, it would be expected that satisfaction surveyed in both IRM-1 and IRM-2. Figure 2.9 com- with service delivery might be impacted. IRM-2 pares responses across the two surveys using the panel survey respondents were asked about their satisfac- data. The greatest rise in dissatisfaction is for water tion level with electricity, drinking water, medical with 21% dissatisfied compared to 11% in IRM-1. The facilities, schools, and roads in both waves of the dissatisfaction level has increased by three percentage survey (Figure 2.8).16 Respondents in IRM-2 were points for electricity, three percentage points for med- most satisfied with schools (85%). The greatest level ical facilities, and two percentage points for motorable of dissatisfaction was with two services: drinking wa- roads. This suggests growing frustration as these ter (21% dissatisfied) and electricity (17%). Only 11% services have not fully recovered or improved. In con- expressed dissatisfaction with motorable roads and trast, IRM-2 respondents who were also interviewed 9% with medical facilities. in IRM-1 were more likely to express satisfaction with schools than before.

16 The figure combines very satisfied and somewhat satisfied as satisfied, and very dissatisfied and somewhat dissatisfied as dissatisfied.

24 Aid and Recovery in Post-Earthquake Nepal

Chapter 3. Earthquake Aid

Photo: Amanda Gurung

IRM-1 was conducted while the emergency earthquake had decreased in reach and volume yet satisfaction response was coming to an end and recovery programs with most aid providers had increased. Contention were beginning. The results from IRM-2 reflect the related to damage assessments, and the corresponding changing nature and volumes of aid as relief shifted to targeting of assistance, did, however, emerge in the a focus on winter-related needs and then preparations IRM-2 research. for reconstruction. IRM-2 found that aid of most types

3.1 Aid received since the 2015 monsoon

The number of people receiving aid has While aid is still primarily concentrated in declined in almost all districts. severely hit districts, there is a trend towards decreased coverage in more affected districts Table 3.1 shows the shares of households not receiving and increased coverage in less affected aid, comparing IRM-1 to IRM-2. All districts surveyed districts. in IRM-2, with the exceptions of Lamjung and Syangja, saw a decrease in the proportion of households receiv- In IRM-1, only 1% in severely hit districts had not ing aid. Whereas in IRM-1, 0-3% of people in each of received any aid since the earthquake. But since June the severely hit districts had not received aid, this has 2015, the share has increased to 5%. Similarly, in risen to 2-8% for IRM-2. A very high percentage of crisis hit districts, the proportion of people receiving Solukhumbu households received aid (95% in IRM-2) no aid increased from 40% (IRM-1) to 59% (IRM-2). despite its being in the third impact category. This is in In contrast, aid coverage has increased in hit with stark contrast to Okhaldhunga, which has seen a par- heavy losses districts, with those not receiving aid ticularly large decrease in the share of people receiving decreasing from 31% to 29%. In the hit district, there aid: from 95% in IRM-1 to 58% in IRM-2 despite the has been a larger decline in the proportion of people district being classified in the second impact category. not receiving aid: from 86% to 70%. Aid provision in rural areas has been higher, with only 24% not receiving any aid compared to 67% in urban areas. For most types of aid, a larger proportion of people in rural

25 Earthquake Aid

areas received assistance than in urban areas. This is Table 3.1: Share of people not perhaps unsurprising given the greater impacts from receiving aid – by district impact and dis- the earthquake in rural areas.17 trict (IRM-1/IRM-2 comparison)

The nature of aid has changed as time has IRM-1 IRM-2 passed with cash becoming the most common form of assistance. Severely hit 1% 5% Dhading 3% 8% IRM-1 found that tarps, food, and cash were the Gorkha 3% 7% most common types of aid received in the emergency Nuwakot 1% 2% response phase of earthquake relief. Nearly one year Ramechhap 0% 5% on from the earthquake, the IRM-2 survey shows a Sindhupalchowk 0% 3% major decline in the share of people receiving tarps Crisis hit 40% 59% (from 71% to 47%) and food (from 53% to 37%). Bhaktapur 46% 58% These types of aid are still common, but have been Kathmandu 70% 77% distributed much less than during the initial weeks Okhaldhunga 5% 42% after the earthquake. In contrast, there has been a Hit with heavy losses 31% 29% large increase in the proportion of people receiving Lamjung 63% 53% cash.18 Since the beginning of the 2015 monsoon, cash Solukhumbu 4% 5% has been the dominant form of aid (Figure 3.1). The Hit 86% 70% focus on winter relief can also be seen in Figure 3.1, Syangja 86% 70% with the distribution of blankets nearly doubling and clothes distribution increasing from 1% of people (IRM-1) to 9% (IRM-2). figure 41

Figure 3.1: Share of people receiving different types of aid – all districts (IRM-1/IRM-2 comparison)

71%

64% 60 53%

47%

40 37% 33% 30% 28% 24% 21% 20 18% 14% 11% 9% 10% 7% 6% 4% 4% 4% 2% 1% 1% 0 0% 0% 0% Tent Tarps I Food edial ash all ash gov ash non govBlanket lothes ithen Sanitation o aid

I-1 I-2

Those in severely hit districts were more likely to have everyone in the severely hit districts, and there have received aid in every category, but the pattern of types also been increases in blankets, CGI, clothes, and of aid received is consistent with that for all impacted sanitation packages. districts (Figure 3.2). Cash has been received by almost

17 Seventy-one percent of houses in rural areas were damaged by the 18 IRM-1 did not disaggregate cash between that from the govern- earthquake compared to 38% in urban areas. However, the impacts ment and that from other sources. on livelihoods were greater in urban areas. See Chapter 2.

26 Aid and Recovery in Post-Earthquake Nepal

figure 42 Figure 3.2: Share of people receiving different types of aid – severely hit districts (IRM-1/IRM-2 comparison)

100 96% 93% 88%

65% 65%

56% 50 43% 38%

26% 25 25%

17% 15% 16% 12% 11% 8% 8% 8% 5% 2% 2% 2% 1% 0 0% 0% 0% Tent Tarps I Food edial ash all ash gov ash non govBlanket lothes ithen Sanitation o aid

I-1 I-2

Cash has been much more widely distributed Amounts of cash per person have also increased. The since the 2015 monsoon than before and is average total cash aid received per household that being provided in increasing volumes. received cash aid was NPR 23,066 in IRM-2, almost double the IRM-1 average of NPR 12,179 (Figure 3.3). The share of people who have received cash from The greatest increase has been in the severely hit any source (government or non-government) has districts, with Syangja (the hit district) seeing a fall risen from 28% in the immediate aftermath of the in the average amount of money provided to each earthquake to 64% (for cash from government) recipient. and 11% (from non-government sources) in IRM-2.

Photo: Subhash Lamichhane

27 Earthquake Aid

figure 4 Figure 3.3: Quantities of cash (NPR) among those who received cash, from all sources – by district impact (IRM-1/IRM-2 comparison)

23,066 Total 12,179

7,616 it 11,429

it with heav losses 15,422 13,204

risis hit 21,414 16,961

Severel hit 26,207 11,311

0 ,000 14,000 21,000 2,000

I-2 I-1

Within the severely hit districts, the reach of cash the government, in Gorkha (Table 3.2). In crisis hit aid has risen from 44% in IRM-1 to 94% in IRM- districts, 46% have received cash aid from any source; 2 (inclusive of government and non-government the figures are 54% for hit with heavy losses districts sources). However, there is variance between districts, and 28% for the hit district. with fewer people receiving cash, especially from

Table 3.2: Amount of cash received (NPR) and share who have received cash – by district impact, district, and source (IRM-2)

Amount received (NPR) Proportion receiving cash Cash aid - government Cash aid – non-government Cash aid - total Cash aid - government Cash aid – non-government Cash aid from any source

Severely hit 25,001 11,717 26,207 90% 17% 94% Dhading 24,552 12,565 26,155 93% 14% 93% Gorkha 19,669 12,821 20,126 70% 31% 88% Nuwakot 29,933 7,944 30,745 96% 9% 96% Ramechhap 24,906 10,000 24,906 95% 0% 95% Sindhupalchowk 24,510 11,227 28,540 97% 32% 97% Crisis hit 19,956 13,750 21,414 45% 6% 46% Bhaktapur 24,229 15,365 26,848 50% 14% 51S% Kathmandu 26,778 11,318 27,327 21% 3% 21% Okhaldhunga 14,220 8,111 14,882 65% 2% 65% Hit with heavy losses 14,732 15,450 15,422 52% 5% 54% Lamjung 23,000 14,423 23,670 23% 8% 27% Solukhumbu 12,420 22,125 12,712 82% 1% 82% Hit 8,202 8,389 7,616 24% 5% 28% Syangja 8,202 8,389 7,616 24% 5% 28% All districts 22,005 12,172 23,066 65% 11% 68%

28 Aid and Recovery in Post-Earthquake Nepal

The relatively high level of non-governmental cash in recipient is slightly higher from non-government hit with heavy losses districts is driven by the amounts sources than from the government. Indeed, cash received in Solukhumbu, the district with the largest from non-government sources is highest in the third share of low income households. The most frequently impact category (NPR 15,450), followed by the second cited sources of such aid in Solukhumbu are NGOs, (NPR 13,750). In contrast, the average amount of INGOs, and the Red Cross. cash aid from non-government sources in severely hit districts (NPR 11,717) is lower than the average Government cash has primarily gone to people amount provided to crisis hit and hit with heavy who have received official beneficiary cards losses districts. At lower levels of impact, where the following government damage assessments. reach of all items, including government cash aid, is lower, non-government players are playing a bigger Ninety-two percent of those with a beneficiary card role in relief efforts, albeit with a smaller reach. Non- received cash from the government compared to government providers are targeting people who have 6% who do not have a card. Similarly, 15% of those not received cash from the government. with a card received cash from non-government organizations compared to 1% who do not have a card Despite increases in the number of people (Figure 3.4).19 getting cash, and the rise in volumes of cash per person, it is clear that cash provided thus far is not enough. Figure 3.4: Share of people who have received cash – by whether or not received The sums provided are insufficient for addressing key figure 25 beneficiary cards (IRM-2) needs such as building more robust housing. For this reason, large numbers of people, especially in severely 0 hit districts but also elsewhere, prioritize cash as one of their priority immediate and medium-term needs 71% (Chapter 2). 66% 60 To date, government programs have provided two 55% cash grants. The first was an allocation of NPR 15,000 to affected households to support the construction of temporary shelters. A second program provided NPR 40 38% 10,000 for winter relief aimed at helping people buy materials, including clothes and fuel, to help them withstand the winter months. Both of these cash grants 20 were only to be provided to households categorized as ‘fully damaged’ in the beneficiary lists generated by locally-led damage assessments.20 The cash was 0 distributed through village-level Relief Distribution rban ural Committees (RDCs).

ousing damage ivelihood damage Citizens and local officials interviewed during the qualitative research complained that cash provided as winter relief was too little and arrived too late. In Gorkha district, for example, the total amount needed to distribute cash assistance to all confirmed On average, recipients of cash from the gov- beneficiaries, an amount of NPR 70 million, was ernment receive more than recipients from sent to the district in multiple tranches, which led to non-government providers. delayed distribution for some areas (prioritization was done on the basis of the level of damage). In Those who received cash from the government got, Lisankhu VDC in Sindhupalchowk district, 97 out of on average, NPR 22,005, while the average sum 1,131 households had yet to receive winter relief cash for non-governmental cash is NPR 12,172. In less at the time of research in February 2016 due to delays affected districts, the average sum provided per cash in the disbursement of funds by the district.

19 See Chapter 3.4 for a discussion of the damage assessments and percent who were given a beneficiary card because their house was beneficiary cards. fully damaged received cash from the government. Six percent of 20 The survey found that this was largely the case. Ninety-two those without a card received government cash.

29 Earthquake Aid

Fewer households received food aid in almost There has been limited assistance for live- all districts after the monsoon. lihoods and distribution of reconstruction materials. The largest reductions in food aid coverage were in Dhading (from 93% in IRM-1 to 38% in IRM-2), Gorkha IRM-2 found that 9% of households (14% in severely (from 89% to 50%), Bhaktapur (from 34% to 11%), hit districts) reported receiving farm implements, and and Okhaldhunga (from 34% to just 2%) – Table 3.3. no-one had received support for livestock rearing. Despite the overall decrease in the coverage of food This suggests a lack of attention to livelihoods support aid, those households that did receive assistance were as 77% of respondents farm their own land for their provided a greater volume of food. Those reporting primary livelihood and a further 18% rear livestock as food aid in IRM-2 received an average of 32 days a primary source of income. There has also been very of food compared to an average of 24 days of food limited distribution of reconstruction materials since reported during the IRM-1 survey. the beginning of the 2015 monsoon, with 6% receiving materials (8% in severely hit districts).21

Table 3.3: Share of people who have received Shelter-related aid has focused on providing food aid – by district (IRM-1/IRM-2 comparison) materials to help strengthen temporary shelters. IRM-1 IRM-2 At the time of the IRM-2 field visits, the distribution Severely hit of cash grants from the government for reconstruction Dhading 93% 38% of housing had not begun. A small minority of Gorkha 89% 50% households were rebuilding their houses with their Nuwakot 96% 87% own money and a very small number of houses being Ramechhap 89% 67% rebuilt with some support from local NGOs. Shelter- Sindhupalchowk 100% 83% related aid has remained focused on tarps and CGI. Crisis hit Bhaktapur 34% 11% The proportion of people receiving tarps has declined Kathmandu 9% 9% since the beginning of the 2015 monsoon season (from Okhaldhunga 34% 2% 71% in IRM-1 to 47% in IRM-2; from 96% to 65% in the Hit with heavy losses severely hit districts) – Table 3.4. However, the large Lamjung 6% 15% amount of people still receiving tarps is surprising Solukhumbu 10% 33% given that initial distribution was so widespread and Hit the fact that tarps offer little protection during the Syangja 3% 4% winter. All districts 37% 37% CGI distribution in severely hit districts has risen since IRM-1 but is still low: 38% against 12% in The food security situation appears particu- IRM-1. Indeed, in IRM-2, households are still more larly serious in Okhaldhunga. likely to have received tarps than CGI. Besides Sindhupalchowk, where 82% have received CGI since Only 2% have received food aid in Okhaldhunga since the beginning of the 2015 monsoon, the proportion of June 2015, far lower than any district. Volumes of food people receiving CGI is much lower than the share of aid in Okhaldhunga are also the lowest of any districts people whose houses were damaged. in both IRM-2 (four days of stock, against the average of 32 days) and IRM-1 (five days compared to 24 days). With the exception of Sindhupalchowk, very few This is especially concerning as Okhaldhunga has a people have received reconstruction materials. high share of low income households (54% have a Only 4% of households overall have received them, monthly income below NPR 10,000). Evidence from including 8% in severely hit districts. the five low income districts of Ramechhap, Dhading, Nuwakot, Solukhumbu, and Okhaldhunga suggests that where food aid does not arrive, people are more likely to borrow for food.

21 Data on the distribution of reconstruction material was not from IRM-1 showed that there was next to no distribution of collected in IRM-1, so it not possible to formally make comparisons reconstruction materials (beyond CGI) in the first months after between the two time periods. However, qualitative fieldwork the earthquake.

30 Aid and Recovery in Post-Earthquake Nepal

Table 3.4: Share of people who have received shelter items – by district impact and district (IRM-1/IRM-2 comparison)

IRM-1 IRM-2

Housing Reconstruction Tent Tarps CGI Tent Tarps CGI damage materials Severely hit 94% 2% 96% 12% 2% 65% 38% 8% Dhading 97% 1% 93% 13% 1% 37% 33% 4% Gorkha 89% 1% 96% 13% 0% 53% 27% 4% Nuwakot 97% 6% 91% 10% 5% 72% 39% 1% Ramechhap 90% 1% 100% 6% 1% 82% 9% 1% Sindhupalchowk 97% 0% 99% 20% 3% 82% 82% 29% Crisis hit 49% 2% 48% 3% 1% 13% 3% 0% Bhaktapur 60% 3% 40% 5% 1% 17% 3% 1% Kathmandu 26% 1% 11% 3% 3% 8% 3% 0% Okhaldhunga 60% 2% 95% 0% 0% 16% 1% 0% Hit with heavy losses 44% 0% 66% 3% 1% 66% 14% 2% Lamjung 35% 1% 35% 2% 1% 41% 16% 2% Solukhumbu 52% 0% 93% 3% 1% 92% 12% 3% Hit 21% 1% 11% 1% 0% 14% 2% 0% Syangja 21% 1% 11% 1% 0% 14% 2% 0% All districts 66% 2% 71% 7% 1% 47% 21% 4%

This lack of distribution of more sturdy building which fares worst in terms of shelter provided, are materials means that demand is high. Forty-seven the most likely to prioritize reconstruction materials percent of households in severely hit districts said (40%). Demand for CGI is also high across the severely materials for housing reconstruction were among their hit districts (18% in severely hit districts list it as a top two needs, with people in Nuwakot and Ramechhap top two priority) and some others (Okhaldhunga and most likely to prioritize such materials. Among the Solukhumbu). non-severely hit districts, people in Okhaldhunga,

Table 3.5: Share of shelter items in top two current needs – by district impact and district (IRM-2)

Reconstruction Tent CGI ‘A house’ materials Severely hit 4% 18% 47% 11% Dhading 1% 20% 39% 6% Gorkha 2% 15% 35% 11% Nuwakot 1% 21% 69% 0% Ramechhap 14% 18% 59% 1% Sindhupalchowk 1% 19% 34% 39% Crisis hit 2% 11% 24% 1% Bhaktapur 2% 11% 16% 3% Kathmandu 0% 3% 17% 0% Okhaldhunga 3% 20% 40% 0% Hit with heavy losses 3% 22% 22% 0% Lamjung 1% 11% 23% 0% Solukhumbu 6% 33% 21% 0% Hit 3% 9% 23% 0% Syangja 3% 9% 23% 0% All districts 3% 16% 34% 5%

31 Earthquake Aid

3.2 Providers of aid

The government, INGOs, and NGOs continue 22% report aid from NGOs and 19% from INGOs to provide aid to the most people. (Figure 3.5). Other aid providers, with the exception of the Red Cross, are much less prominent. The same Respondents in both IRM-1 and IRM-2 were asked to groups are the main sources of aid in the severely provide information on the sources from whom they hit districts, although they cover larger shares of the received aid.22 The proportion of surveyed household population than in other districts. reporting aid from the government is 66%, while

figure 46 Figure 3.5: Source of aid – all districts (IRM-1/IRM-2 comparison)

0 66%

61%

22% 19%19% 1 14% 11% 11% 9% 4% 3% 4% 3% 2% 1% 1% 1% 1% 1% 1% 1% 1% 1% 0 1% ov itiens Armpolie arties ed ross Individuals Business I ountries onors

I-1 I-2

Government aid activity decreased after NGO and INGO aid has also reached a larger the monsoon and was overwhelmingly number of people in severely hit districts than focused on the cash assistance schemes for in others but there has been a sharp rise in aid the construction of temporary shelters and coverage in hit with heavy losses districts. winter relief. Aid coverage in severely hit districts by NGOs has Contributing factors to these trends included the risen from 21% in IRM-1 to 33% in IRM-2. There has formal end of the emergency relief period (as declared also been a large increase in the proportion of people by the government in June 2015), and the national who are receiving such assistance in the hit with heavy political focus on the new constitution along with losses districts (Figure 3.7). efforts to address the unrest in the Terai. Despite this, the household survey found that 90% of people in Nineteen percent of households in both IRM-1 the most affected districts had received government and IRM-2 reported receiving aid from INGOs support in IRM-2 (Figure 3.6). (Figure 3.8), with coverage rising in hit with heavy losses districts. This increase is driven primarily by aid provided to Solukhumbu, where 35% have received aid from INGOs in IRM-2 compared to 4% in similarly 22 It should be noted that responses are based on the perceptions or understanding of respondents, so results cannot be taken affected Lamjung district. There was a slight reduction to necessarily reflect the actual shares of aid being provided by in aid from INGOs in severely hit districts and almost different providers. no change in crisis hit areas.

32 Aid and Recovery in Post-Earthquake Nepal

figure 4

Figure 3.6: Share of people receiving government aid – by district impact (IRM-1/IRM-2 comparison)

0 90%

82%

6 66% 61% 57%

47% 48% 45 45%

29% 2

13%

0 Severel hit risis hit it with heav losses it Total

I-1 I-2

figure 4

Figure 3.7: Share of people receiving NGO aid – by district impact (IRM-1/IRM-2 comparison)

33%

24% 21% 22% 20

14%

10 8% 7% 7%

1% 1% 0 Severel hit risis hit it with heav losses it Total

I-1 I-2

33 Earthquake Aid

figure 410

Figure 3.8: Share of people receiving INGO aid – by district impact (IRM-1/IRM-2 comparison)

34%

31% 0

20 19% 19% 19%

6% 6% 5%

0 0% 0% Severel hit risis hit it with heav losses it Total

I-1 I-2

NGOs and foreign government development activities. Some foreign development agencies also agencies have provided a more diverse range focused reconstruction support on government of aid than government, with a focus on non- and community buildings (especially in health and food relief schemes. education) and physical infrastructure (roads and bridges). In the three severely hit districts studied, Field research identified recovery-related activities organizations including UNICEF, Care Nepal, and of a number of types: cash grants, sometimes as top- Plan International, provided cash, warm clothes, ups to government cash assistance, through direct and other winter relief materials to complement the distribution as well as cash for work schemes (such as government’s winter relief cash distribution and other rubble removal, road clearing, and debris management cash assistance schemes. Winterization relief was seen schemes); construction materials for private and as a success by many international actors as there community temporary shelters; a limited amount were no large outbreaks of disease and monitoring of food relief provided directly or through food for mechanisms involving the government and the UN work programs; and masonry and carpentry training reportedly worked well. for village residents in anticipation of reconstruction

3.3 Coordination, accountability, and transparency

IRM-1 found that while aid distribution was The coordinating committees that were active initially chaotic, coordination improved after in the emergency relief phase have become the formalization of government mechanisms. less active in the months after the monsoon.

Relief committees were established at the district DDRCs have largely become inactive but other and VDC levels: District Disaster Relief Committees local actors have bridged the gaps and maintained (DDRCs) and VDC relief distribution committees coordination efforts. Most district level stakeholders (RDCs). One function was to coordinate all aid coming in, both from the government and from non- governmental providers. This ‘one door’ policy was 23 The Asia Foundation (2005). Independent Impacts and Recovery seen to have been effective and the DDRCs and RDCs Monitoring Phase 1: June 2015. Synthesis Report. Kathmandu and were perceived to have performed well.23 Bangkok: The Asia Foundation, pp. 24-26.

34 Aid and Recovery in Post-Earthquake Nepal

interviewed reported that their district’s DDRC now Lack of coordination between different levels meets “only when necessary,” which seems to mean of government has meant that both local when new relief schemes arrive. Different districts officials and earthquake affected people do have seen different combinations of actors filling the not have the information needed to help them gap caused by a less active DDRC. recover.

At the VDC level, the main coordination challenge In the absence of clear directives and information on is weak communication with the district and the government policies, VDC level officials and citizens central government. Communication between district report that they have to rely on informal conver- headquarters and the VDCs, which research found to sations, direct phone calls, and media reports for have mostly taken place through the VDC secretaries, information on reconstruction policies. The lack of is happening in an informal and unstructured way. information that most local government officials have Communication approaches differ between VDCs, and has made them quite passive, waiting for instructions generally the situation does not lend itself to uniform from the central government before taking action. coordination. While VDC level actors have remained People do not understand whether they will be entitled central to managing and coordinating aid delivery, as to reconstruction assistance or not, when support for well as dealing with pressure from citizens, the local this may arrive, and what procedures there are for coordinating body, the RDCs, were also found to be making complaints. This is hampering their ability less active than they had been during the initial relief to plan for the future – e.g. on whether they should period. start rebuilding their house, or whether they need to wait, and on whether or not they should take out loans (Case Study 4).

Case Study 4: Indecision due to unclear and delayed government policy A 57-year-old resident of Katunje-4 had been headquarter to gather information. Once there, living in a hut made of tarpaulin and dried someone showed him a few models of quake- tree-leaves (shyaula) for nine months. After resistant houses that should guide rebuilding. he received a beneficiary card, and NPR 15,000 He quickly realized that he could not afford to from the government, he demolished his dam- build according to the model house standard. aged house. Unclear on government policy and He is currently planning to construct a small guidelines for rebuilding houses, he found it dif- building which will be stronger than his old ficult to decide whether or not to start building a house saying that he “want[s] to build a house new house. After discussion with his family, he that does not leak in the monsoon and can be ultimately decided to construct a new house so easily abandoned if need be.” Despite this plan, that his family would not have to suffer another he has delayed the start of construction until he cold winter and wet monsoon in the leaf hut. receives more information.

He started building the new house. He reported In another case, a woman in Okhaldhunga that he “carried stones from a place that was a district who had been living in a bamboo four hour walk away and paid more than NPR shed had bought gravel and sand to start the 10,000 for the stones. I also paid for laborers to construction of a new house. But other villagers dig a six-foot deep trench for a strong foundation advised her that building a house which does not to start building a two-room structure.” Once he comply with the model set by the government had started building the house, other villagers may bar her from getting government financial told him about a rumor that starting to build assistance for rebuilding. She told researchers without the permission of government officers that this possibility has left her confused and would bar him from getting any compensation that she has not been able to decide to start the and financial assistance from the government construction of a new house. She says that if in the future. Some village leaders asked him to “the government informs us about its assistance stop the construction, and he complied. and the criteria, I would start to buy rebuilding materials. But I will not unless the government To try to clarify the matter, he undertook the two will give us clear guidance.” hour walk and four hour bus ride to the district

35 Earthquake Aid

Coordination of I/NGOs, foreign agencies, The majority of NGOs reported a preference for working and UN agencies with local governments has through their own staff for targeting and distribution been less effective in recent months than it with direct communication with earthquake-affected was during emergency relief distribution. persons (sometimes through WCFs). NGO aid is now more likely to be targeted to specific population Increasingly, non-governmental aid has bypassed offi- groups. This is a result in part of overall declines in aid cial coordination mechanisms and is working directly since the initial relief phase. Because of sensitivities in with citizens or ward level actors. IRM-2 qualitative targeting aid within communities where earthquake interviews found that village level RDCs were the most impacts are widespread, it is problematic that VDC common channel for the distribution of government stakeholders have been less involved in deciding who relief. On the other hand, most non-governmental should receive aid. and individual aid providers worked by reaching out directly to citizens. Non-governmental aid providers There has been extremely limited citizen were also more likely than before to work with Ward participation in targeting or overseeing gov- Citizen Forums (WCFs) and ward leaders directly to ernment aid distribution mechanisms, par- help coordinate and target the distribution of relief. ticularly from marginalized groups.

Initially strong coordination efforts suffered for a By design, local communities were represented at RDC number of reasons including: the reduced DDRC meetings by WCF coordinators. However, their role activity; local government liaison officers returning to was limited. These individuals played an important their former posts; local government staff turnover; role in implementing aid distribution, but they had no and humanitarian staff turnover. In addition, the influence over planning and decision-making, which withdrawal of UN OCHA field offices (OCHA left Nepal were led exclusively by VDC officials and political in December 2015), and the closing or transition of parties. VDC officials often emphasized that they had most humanitarian clusters in late 2015, has affected ensured citizen participation in relief efforts through coordination. the inclusion of existing WCFs, and adhering to the statutory policy of those bodies to include women and The lack of coordination forums has hindered the marginalized groups in decision-making. However, the efforts of District Lead Support Agencies (DLSAs, majority of individuals interviewed during the field typically I/NGOs who are assigned to support disaster research who belonged to these particular groups per- preparedness and response) to support coordination ceived these efforts as superficial saying that inclusion between aid providers and local government. OCHA was “cosmetic” or “token”, only done to “fill the quota”. field offices, based in the three humanitarian hubs (Gorkha, Sindhupalchowk, and Kathmandu), had been Both governmental and non-governmental able to take a cross-district perspective in managing aid distribution lacked formal accountability these efforts, something which DLSAs cannot do. mechanisms.

Non-governmental organizations and foreign govern- Accepted good practice to ensure transparency and in- ment development agencies generally just informed crease accountability during local development plan- DDRCs and bypassed VDC level committees when ning and implementation includes needs assessments distributing aid. The levels of coordination between and planning sessions conducted at the ward level, non-governmental organizations and village level formation of a monitoring committee separate from RDCs were inconsistent across the VDCs visited.24 the project implementing committee, and a public or Non-governmental aid frequently bypasses the RDCs, social audit (which is a public display/announcement despite their formal role as the focal point for the of project income/expenses). However, researchers planning and coordination of aid distribution. In most only came across one public audit of relief and recov- cases, there was little or no coordination regarding ery projects, which was attended mainly by the heads actual distribution and the process for targeting aid of the government agencies, with very limited citizen to certain beneficiaries. During IRM-2, local officials participation. The discussion was mainly focused on often had trouble reporting what aid had been received the accomplishments of district-level government in their area. agencies and their plans for the future, rather than on earthquake-related aid distribution and reconstruction planning.

24 It is important also to consider that NGOs have reasons to be wary of local government involvement, which may translate into directing aid based on unreliable data and political interference.

36 Aid and Recovery in Post-Earthquake Nepal

No clear evidence of widespread abuse of funds then VDC levels, and was handled by a few influential or goods was found. But the ineffectiveness individuals at each stage of the process. In severely of accountability mechanisms suggests that hit districts, such as Gorkha and Sindhupalchowk, corruption related to aid distribution is likely the cash amounts given by the central government to to have been underreported. the districts were in excess of NPR 750 million. Since these large cash volumes were distributed on the Only a few clear instances of corruption were reported basis of widely contested damage assessments, there across the six research districts during qualitative is a possibility that the assessment might have led to field research. The government’s aid distribution corrupt practices in some places. The government methods, which have overwhelmingly focused on the reported problems counting households, as figures direct distribution of large amounts of cash, make it for the total number of households in each district particularly vulnerable to some forms of corruption. substantially increased compared to the 2011 census, This cash moved from the center to the district and again raising the possibility of corruption.

3.4 Damage assessments and beneficiary cards

The government has conducted a number of assess- Many complaints were due to lack of trans- ments of damages to houses, using these as a basis for parency and consistency on how houses were deciding who will receive a beneficiary card.25 These classified in the damage assessment and how cards determine who should receive cash assistance, this links to the issuing of beneficiary cards. including the two payments made over the winter and a larger future grant of NPR 200,000 to support Table 3.6 compares the results of the damage assess- housing reconstruction. Only those whose house was ment with households’ self-reporting of the extent to classified as fully damaged were to receive cards.26 which their house was damaged. In general, there is a close correlation between being classified as fully The assessments have been a cause for reg- damaged in the assessment and households reporting ular complaints from earthquake-affected that their house was completely damaged. In some households. districts, such as Gorkha and Ramechhap, more people had their houses classified as fully damaged than During the qualitative research residents cited mul- report their house as being fully damaged. In contrast, tiple reasons for frustration with the damage assess- 53% of people in Dhading say their house was classi- ment process including: inconsistent assessment fied in the assessment as fully damaged while 77% say procedures; a lack of clear policies and guidance; that their house was completely damaged. assessment teams without technical knowledge; differences between the multiple assessments that took place; and perceived manipulation and interference by political parties.

25 VDCs completed their first damage assessment within a week of beneficiary list for each VDC. This DDRC-led assessment focused on the 25 April earthquake, usually using local actors (VDC officials, damage to houses, categorizing each house into three levels: ‘fully local political party leaders, schoolteachers, and WCF members) damaged’, ‘partially damaged,’ and ‘normal’. Households found to to complete the data collection. These initial assessments were have houses that were fully damaged were given victim ID cards informal and ad hoc in nature, but were significant in helping to and prioritized for government relief. As mentioned in the first manage initial emergency relief. After the DDRCs formed in the round report, because of the direct link of the assessment to aid weeks following the first earthquake, they conducted a second provision, and the seemingly disorganized way it was conducted damage assessment, which was generally carried out within a month. in many locations, this second assessment became a significant Whereas the initial assessment aimed only to support emergency source of contention throughout all districts. The Asia Foundation relief distribution, the second assessment was an attempt to gather and Democracy Resource Center Nepal (2015). Aid and Recovery more comprehensive data. This second assessment was supposed in Post-Earthquake Nepal: Independent Impacts and Recovery to be carried out by a team that included an engineer. While it Monitoring Nepal Phase 1 – Qualitative Field Monitoring (June was implemented by the DDRC, instructions were provided by the 2015). Kathmandu and Bangkok: The Asia Foundation. central government on how it should be conducted. The second 26 See Chapter 3.4 of the qualitative report for extensive analysis of damage assessment was used to generate an earthquake relief the damage assessments.

37 Earthquake Aid

Table 3.6: Damage assessment results and self-reported damage – by district impact and district (IRM-2)

Damage assessment results Self-reported damage Fully damaged Partially damaged Normal/not damaged Official did not arrive Refused/don’t know Completely damaged Badly damaged (needs major repairs) Habitable (but needs minor repairs)

Severely hit 80% 6% 2% 6% 6% 78% 16% 5% Dhading 53% 5% 3% 28% 11% 77% 20% 3% Gorkha 80% 13% 6% 0% 2% 57% 32% 8% Nuwakot 92% 3% 1% 2% 3% 92% 5% 3% Ramechhap 83% 6% 1% 1% 10% 69% 20% 10% Sindhupalchowk 92% 3% 1% 1% 4% 94% 3% 2% Crisis hit 36% 21% 36% 2% 5% 31% 18% 30% Bhaktapur 50% 11% 31% 1% 7% 46% 14% 15% Kathmandu 16% 16% 59% 3% 6% 16% 10% 38% Okhaldhunga 43% 37% 19% 0% 2% 32% 29% 37% Hit with heavy losses 31% 39% 22% 0% 8% 27% 17% 42% Lamjung 27% 19% 38% 0% 16% 24% 11% 39% Solukhumbu 35% 58% 6% 0% 1% 29% 23% 45% Hit 11% 22% 66% 0% 1% 8% 13% 62% Syangja 11% 22% 66% 0% 1% 8% 13% 62% All districts 53% 17% 21% 3% 6% 50% 16% 24%

There is also variation between districts in how damaged, 93% received beneficiary cards, higher damage assessment classifications resulted in cards even that people’s self-evaluation of the impact of the being issued. In Dhading, for example, where 53% earthquake on their house (Figure 3.9). of people said their house was classified as fully

figure 41 Figure 3.9: Share of people who have received beneficiary card and whose house was classified as fully damaged – by district (IRM-2)

100 99% 96% 94% 93% 92% 92% 89% 85% 83% 80% 5

60%

53% 51% 50 50% 43%

35% 35% 29% 27% 25 22% 16% 11%

0 hading orkha uwakot amehhap SindhupalhowkBhaktapur athmandu khaldhunga amung Solukhumbu Sanga

eeived benefiiar ard Full damaged in damage assessment

38 Aid and Recovery in Post-Earthquake Nepal

Rumors over the implications of the classifi- cards actually accepted and received the initial cash cation also affected household decisions on grant of NPR 15,000. applying for beneficiary cards. In Ramechhap district, some victims applied to change In the absence of clear information and guidelines, the categorization of their damaged houses from fully misinformation was common and rumors affected damaged to partially damaged following a rumor the decisions that households made in many cases. In that fully damaged houses would be demolished by Syangja district, for example, the CDO reportedly told the Nepal Army. Another rumor in Sindhupalchowk people that they should return their beneficiary cards district spread false information that the family if they did not want their houses to be demolished. members of those with beneficiary cards would get As a result, in Syangja’s Shreekrishna Gandaki VDC, opportunities for foreign university scholarships. It six households returned their victim ID cards and was reported that as a result some people permanently received no cash assistance. For the same reason living in Kathmandu returned to their home villages in Waling Municipality, Syangja district, only 18 and applied for cards, thus inflating the official number households out of 250 who had originally received of affected persons in that area.

Case Study 5: Rumors A female owner of a small hotel in Nele VDC-7 Fearing that her house would be demolished, in Solukhumbu lived in a tent for two months and that the reconstruction grant would not after seeing major cracks in the walls and ceil- be sufficient to pay for its reconstruction, she ings of her house following the earthquake. lobbied VDC officials to have the classification When the second earthquake hit on 12 May, her of her house changed from fully damaged to house was further damaged. A local assessment partially damaged. No one questioned her team, comprised of a police officer, a teacher, decision and her status was changed. By the and a WCF coordinator, categorized her prop- time she found out that it had been just a rumor, erty as fully damaged and she received a victim the inspector who had given her the information beneficiary card. Before the distribution of cash had been transferred to another location, and grants started, a rumor spread in Nele that the VDC Secretary stated that no change could houses categorized as fully damaged would be be made since she had already accepted a cash demolished by the Nepal Army before recon- grant of NPR 3,000 provided for those with struction would start. The women claimed that partially damaged houses in June. she heard this rumor confirmed by an inspector posted to the ward.

39 Earthquake Aid

Those who received beneficiary cards tended are somewhat satisfied –Table 3.7. In contrast, there to be satisfied with the classification of their are mixed levels of satisfaction amongst those who house in the damage assessment. did not receive cards. Two-thirds of this group are still satisfied, but one-quarter say they are unsatisfied. Sixty-one percent of those who received cards are very satisfied with the damage assessment and 27%

Table 3.7: Satisfaction with damage assessment – by whether or not received beneficiary cards (IRM-2)

Has your household received a beneficiary identity card? Total Yes No Refused Don’t know

Very satisfied 61% 22% 0% 0% 49% How satisfied were Somewhat satisfied 27% 43% 100% 23% 32% you with the classifi- Somewhat unsatisfied 3% 15% 0% 8% 7% cation of your house Very unsatisfied 1% 10% 0% 19% 4% in the official dam- Refused 0% 1% 0% 4%% 0% age assessment? Don’t know 7% 9% 0% 46% 8% Total 100% 100% 100% 100% 100%

The frustration of those who did not receive a card, levels are the lowest for those whose house was even though their houses were damaged to some partially damaged, presumably because many felt they extent, can be seen in the IRM-2 household survey should have received assistance but have not been results. Of those whose house was classified as fully issued with beneficiary cards that help them access damaged, almost everyone is satisfied. Satisfaction aid (Table 3.8).

Table 3.8: Satisfaction with damage assessment – by how house was classified in damage assessment (IRM-2)

How satisfied were you with the classification of your house in the official damage assessment?

Total Very Very satisfied Somewhat satisfied Somewhat unsatisfied Very unsatisfied Refused know Don’t

Fully damaged 73% 25% 1% 0% 0% 1% 100% Partially damaged 28% 44% 16% 11% 0% 1% 100% How was your house Normal/not damaged 26% 50% 15% 6% 0% 3% 100% classified in the official Official did not arrive 3% 9% 2% 5% 1% 80% 100% damage assessment? Refused 0% 33% 67% 0% 0% 0% 100% Don’t know 1% 5% 8% 10% 2% 74% 100% Total 49% 32% 7% 4% 0% 8% 100%

However, levels of satisfaction with the satisfied with how their house was classified, 58% damage assessment are not fully determined received cards while 42% did not. Furthermore, by whether or not people received beneficiary substantial shares of those who received beneficiary cards. cards were not satisfied with how their house was classified. This suggests that some people were not Of people who are very satisfied, 86% received happy with the damage assessment process, even beneficiary cards. Amongst those who are somewhat when they received a beneficiary card.

40 Aid and Recovery in Post-Earthquake Nepal

Photo: Alok Pokharel

Complaints increased when people realized DDRC created a technical team in December 2015 to the implications of how their house was investigate 6,181 complaints about the results of dam- classified in the second damage assessment. age assessments. In Sindhupalchowk district, almost The ways in which DDRCs handled these 3,400 official objections were made about households complaints varied from district to district. being wrongly included in the beneficiary list (which had been based on the very first damage assessment). Officials received many claims and complaints The DDRC formed a monitoring committee to address regarding the outcome of the second assessment. the complaints, but a team from the Central Bureau These increased when it became apparent that the of Statistics (CBS) arrived in the district to conduct a government would only provide relief for building third damage assessment while the complaints were transitional shelters to beneficiary card holders, and still being assessed so the effort was discontinued. that only those with fully damaged homes were eligible for the cards. At this point citizens started to lodge Contention around damage assessments and the re- complaints claiming that they had been left out of sulting beneficiary lists still exist. When the NRA was the assessment, or that their house had been wrongly formed, its leadership decided that due to the highly categorized. In most cases, citizens lodged their contested nature of the initial damage assessments, complaints at the VDC office, which were generally and to respond to donor requests, another damage then forwarded to the DDRC in almost all locations. assessment was needed. This NRA-led assessment was conducted using CBS enumerators. This decision In most districts, the DDRC investigated the com- has been controversial and the new assessments have plaints, and often added some households to the bene- already been contested during the NRA’s first effort ficiary list as a result. Several districts received a large to conclude reconstruction grant agreements with af- number of complaints. In Gorkha, for example, 59,523 fected households in Dolakha. Many people protested households received victim ID cards after the second against distribution on the basis of the CBS verified damage assessment. After thousands of complaints data. Instead they demanded that the distribution be were subsequently lodged, the district formed a team based on the original—second—assessment, which comprised of the District Lawyer and representatives seems to have reported many more ‘fully damaged’ from the District Administration Office to address the houses than were identified in the initial phases of complaints. By the time of the field research, this team the NRA assessments. Frustration with the newest had increased the number of victim ID cards issued damage assessment is the latest in a series of disputes to 66,144, but thousands of complaints in Gorkha that have arisen from the process of assessing damage remained outstanding.27 In Okhaldhunga district, the across multiple rounds of assessments.

27 It should be noted that this increase by the redress committee census of 2011. CBS, National Population Census 2011, November meant that the number of households that were included in the 2012 (available at: http://cbs.gov.np/image/data/Population/ beneficiary list (66,144) ended up being higher than the total Ward%20Level/36Gorkha_WardLevel.pdf). number of households in Gorkha district according to the national

41 Earthquake Aid

3.5 Satisfaction with aid providers

Despite problems with damage assessments, but they still have the highest satisfaction levels among declining levels of aid, and lack of accountabil- any aid provider. The army and police have actually ity mechanisms, there has been a rise in satis- delivered very little assistance: only 1% had received faction with most aid providers since IRM-1. aid in both IRM-1 and IRM-2 from them. People may still have positive memories of the rescue role they This increase has been greatest for local community played in the days following the earthquake. Satisfac- organizations (Figure 3.10). The army, police, and tion with both INGOs and NGOs has increased and is businesses have seen the greatest drops in satisfaction, similar to levels of satisfaction with the government.

figure 411 Figure 3.10: Satisfaction with aid providers – all districts, among those who received aid only (IRM-1/IRM-2 comparison)

0 85% 83%

76% 74% 71% 69% 6 66% 66% 66% 62% 59% 57% 59% 59% 54% 50% 47% 45 39% 39% 33% 32% 30%

0 ov Arm olie olitial partiesAdmin Is s Business ommunit groupsForeign governmenteligious groups

I-1 I-2

People are most satisfied with aid providers in that more aid has flowed to rural areas. On the other, the hit with heavy losses districts and urban needs are also higher in these places. Dissatisfaction areas. with the government amongst those who received aid is notably higher in urban areas than in rural ones and Solukhumbu has the highest satisfaction of any district there is also much lower satisfaction with political for nearly every aid provider, which might be related parties in urban places. to its receiving very high aid coverage compared to the level of damage suffered (Table 3.9). Okhaldhunga, on While the overall trend is of increased sat- the other hand, has the lowest satisfaction with most isfaction with aid, the qualitative research aid providers among all districts. Aid has fallen sharply found frustration among many communities in this district in IRM-2, particularly for essential with the aid delivered. items (food and shelter). However, even those who receive aid are less satisfied there, perhaps reflecting In many locations studied there was contention around that either volumes have not been sufficient or that cash assistance, often related to the problematic there is disillusionment as other needy people in the damage assessments on which the RDCs based the district miss out. grant distribution. Many earthquake-affected citizens noted their extreme dissatisfaction about being Satisfaction with most aid providers is higher in rural excluded from the cash beneficiary lists (complaints areas. On the one hand, this is not surprising, given came both from cases in which a fully damaged house

42 Aid and Recovery in Post-Earthquake Nepal

Table 3.9: Satisfaction with aid providers, among those who received aid only – by district impact and district (IRM-2)

Districts Government Army Police Political parties Local administration INGOs NGOs Businesses Community groups Foreign governments Religious groups

Severely hit 66% 75% 74% 31% 57% 72% 67% 37% 58% 56% 47% Dhading 70% 83% 82% 20% 49% 60% 64% 31% 52% 31% 34% Gorkha 53% 78% 84% 26% 67% 88% 74% 55% 72% 75% 45% Nuwakot 45% 70% 70% 26% 42% 79% 81% 41% 66% 56% 74% Ramechhap 83% 77% 75% 49% 57% 57% 55% 30% 53% 56% 34% Sindhupalchowk 69% 65% 56% 33% 68% 81% 64% 31% 48% 65% 56% Crisis hit 56% 66% 62% 22% 55% 50% 44% 26% 47% 41% 32% Bhaktapur 49% 66% 63% 20 % 35% 46% 35 % 26 % 57% 37% 43 % Kathmandu 44% 100% 89% 33% 61% 67% 67% 61% 67% 72% 61% Okhaldhunga 63% 58% 55% 20% 69% 49% 45% 19% 36% 38% 18% Hit with heavy losses 81% 93% 93% 49% 71% 92% 92% 62% 81% 93% 64% Lamjung 71% 93% 93% 36% 64% 89% 93% 57% 80% 93% 61% Solukhumbu 94% 94% 94% 69% 81% 95% 92% 69% 81% 94% 67% Hit 76% 89% 92% 59% 76% 78% 78% 70% 89% 86% 57% Syangja 76% 89% 92% 59% 76% 78% 78% 70% 89% 86% 57% All districts 66% 76% 74% 33% 59% 71% 66% 39% 59% 59% 47% was wrongly labeled as only partially damaged, or researchers were told that government and non-gov- from households that were missed by the assessment ernment aid providers still preferred equal distribu- altogether). tion, but to a lesser extent compared to initial relief phase due to decreases in amounts of relief available. Another issue that drove dissatisfaction in some The data collected in this round of research show that places was the local targeting strategy for aid. Equal almost half of the government schemes were targeted distribution of aid was the preferred targeting pattern towards a specific group of beneficiaries (for example, for most aid schemes encountered, a strategy used to reconstruction and winter cash assistance were pro- avoid potential tensions that could arise from per- vided only to fully damaged households). This change ceived discrimination. This was particularly the case in targeting was welcomed by some, particularly those in severely hit districts where levels of damage were disadvantaged communities that felt they had greater largely uniform and relief amounts were sufficient for need than high caste or wealthy community members, equal distribution. During the IRM-2 data collection, but it also caused some frustration.

Photo: Ashray Pande

43 Earthquake Aid

Photo: Amanda Gurung

3.6 Aid to different population groups

The poorest and the richest are the most likely by the earthquake.29 The low figures for the poorest to have received no aid. can be attributed in part to housing structures. The qualitative research found that many of the poorest Disaggregating by types of aid received, the richest families lived in bamboo houses or shacks which were and poorest are the least likely to have received almost not damaged by the earthquake. The targeting of aid every kind of assistance (Table 3.10). This is true for to those whose houses were damaged in some cases aid from the three main providers: the government, limited the ability of the poorest to access those types NGOs, and INGOs.28 That the poor are missing out of assistance.30 However, satisfaction with most aid is worrying, given that the poorest were amongst providers is higher for those in the lowest income the most likely to have been negatively impacted bracket and lower for the richest.31

Table 3.10: Types of aid received – by income band (IRM-2)

Monthly income Tent Tarps CGI Food Medical Cash - government Cash – non- government Blanket Clothes Kitchen items Sanitation No Aid

NPR 40,000 5% 28% 6% 10% 1% 31% 9% 19% 6% 1% 7% 60% Total 1% 47% 21% 37% 4% 64% 11% 33% 9% 4% 14% 30%

28 Sixty percent of those who had a monthly income of under NPR 29 IRM-1 survey report, pp. 10-11. 2,500 before the earthquake who received aid, received it from 30 Case Study 6.3 in the IRM-2 qualitative report. the government, compared to an average of 66% across the whole 31 sample. The figures are 14% compared to an average of 22% for Table 4.19 in the IRM-2 survey report. NGOs, and 7% compared to an average of 19% for INGOs.

44 Aid and Recovery in Post-Earthquake Nepal

Women were slightly less likely to get most Janajatis were more likely to have received types of aid but there is little difference in aid than low or high caste people but low caste satisfaction levels. people were more satisfied with aid providers.

Thirty-one percent of women report receiving no aid Janajatis are particularly more likely to have received compared to 29% of men. Although the differences CGI, food, and government cash (Table 3.11). There are minor, women have a slightly lower likelihood of is little difference between the groups for non- receiving cash from government (62% versus 65%) and government cash and lower caste people are almost non-government sources (10% versus 11%). A slightly as likely to have received tarps as Janajatis. Despite larger proportion of women than men have received being less likely to receive aid than Janajatis, low caste food aid (37% against 36%) and clothes (10% against people are much more likely to be satisfied with aid 9%). Women are slightly more likely than men to be providers than others (Figure 3.11). satisfied with the three main aid providers.

Table 3.11: Share of people who have received aid of different types – by caste groups (IRM-2) Tent Tarps CGI Food Medical Cash - government Cash – non- government Blanket Clothes Kitchen items Sanitation No aid

Low 1% 49% 19% 31% 4% 57% 12% 33% 9% 6% 12% 35% Janajatis 1% 50% 25% 42% 5% 68% 11% 36% 9% 5% 16% 26% High 1% 42% 14% 28% 2% 59% 10% 27% 9% 3% 11% 36%

Figure 3.11: Satisfaction with aid providers among those who received aid – by caste (IRM-2)

41% eligious groups 50% 50% 53% Foreign government 61% 69% 60% ommunit groups 60% 57% 39% Business 40% 36% 62% s 68% 70% 66% Is 73% 80% 54% oal administration 60% 70% 30% olitial parties 34% 38% 78% olie 73% 74% 78% Arm 75% 77% 66% overnment 65% 73% 0 20 40 60 0

igh Janaati ow

45 Earthquake Aid

Those with disabilities were slightly more related assistance, especially tarps, reached those who likely to have received aid than those without, did not experience any housing impact (Table 3.12). although the differences are not large. The likelihood of receiving cash is the highest amongst those whose houses were completely damaged and This holds true for most of the main types of aid. declines as housing impact reduces. Overall, less Across all the main aid providers, a smaller share of than 5% of those with completely damaged homes those with disabilities report being satisfied relative report not having received cash aid from any source. to those without disabilities. It is more concerning that over one-quarter of people with badly damaged homes have not received cash aid Those whose houses were destroyed were the from any source. Two-thirds of those with damaged most likely to have received aid but there is but habitable homes have not received cash. The also some evidence of mistargeting. volumes of cash received for each beneficiary do not correlate closely with housing impact, likely because People whose houses were destroyed were more likely government cash grants did not vary significantly not only to receive shelter-related aid but also other among those who qualified for a given assistance types of assistance. A small proportion of shelter- program.

Table 3.12: Aid received – by level of housing damage (IRM-2) Tent Tarps CGI Food Medical Cash - government Cash – non- government Blankets Clothes Kitchen items Sanitation No Aid

Completely damaged 2% 63% 36% 58% 7% 91% 18% 54% 13% 6% 21% 6% Badly damaged 1% 48% 15% 30% 3% 67% 8% 26% 13% 7% 15% 22% Habitable 0% 30% 2% 12% 0% 32% 2% 7% 2% 0% 4% 57% Not damaged 0% 4% 1% 2% 0% 2% 0% 2% 1% 0% 1% 94%

46 Aid and Recovery in Post-Earthquake Nepal

Chapter 4. Coping Strategies

Photo: Chiran Manandhar

Given the significant damage suffered across earth- more wage labour and communities develoing labor quake-impacted areas, and the limitations in aid rotations to construct temporary shelter. Many of volumes and coverage, households have had to adopt these were unique or temporary measures. Four of the various coping strategies to help them recover. A most common coping strategies are borrowing, inward variety of coping strategies were found during the remittances, migration, and asset sales. qualitative research, including farmers taking on

4.1 Borrowing

There has been a large increase in borrowing Of the 42% of respondents reporting borrowing in since IRM-1. IRM-2, the largest share (60%) borrowed to support their livelihoods, which typically refers to the repair Qualitative researchers found that residents in half of and replacement of damaged assets (Figure 3.2). This the 36 wards visited reported increased borrowing in is true for all four categories of district impact. Seven- their communities. The household survey found that teen percent borrowed for farming or business inputs, the number of people borrowing has more than dou- investments also related to livelihoods. Borrowing to bled since the immediate post-earthquake period purchase food was the second most common reason (Figure 4.1). Whereas in June 2015, 19% had borrowed (35% of borrowers took loans for food), across all lev- since the earthquake, 42% report taking loans since els of impact, including in the least affected district of the beginning of the 2015 monsoon (June 2015). In Syangja. The fact that borrowing for food has been so severely hit districts, more than half have borrowed high in Syangja is surprising given that only 10% in money since the beginning of the 2015 monsoon com- that district report that sources of income have been pared to one-quarter in the first few months after the affected, with the share being slightly less (9%) for ag- disaster. Borrowing has also more than doubled in the riculture on own farms, the occupation of over 90% in crisis hit and hit with heavy impact districts. In the hit the district. One-quarter of people who borrowed did district of Syangja (the least affected of the sampled so to rebuild or repair their home (24% in severely hit districts), borrowing has risen ten-fold: from 4% to 43%. districts) and 20% to finance temporary shelter (31% in severely hit districts). Borrowing is most common for livelihoods, food, and shelter.

47 Coping Strategies

figure 1

Figure 4.1: Share of people who have borrowed since June 2015 – by district impact (IRM-1/IRM-2 comparison)

42% Total 19%

43% it 4%

25% it with heav losses 11%

37% risis hit 17%

52% Severel hit 25%

0 15 0 45 60

I-2 I-1

Figure 4.2: Reasons for borrowing, share of those borrowing – by district impact (IRM-2)

0 65% 62% 60%

5 40% 39% 37% 35% 5 31% 30% 30% 25% 24% 23% 21% 18% 1 18% 9% 9% 7% 6% 5% 5% 4% 3% 3% 2% 1% 1% 1% 1% 1% 1% 1% 1% 0% 0 0% ivelihoods Food Farming Temporar ebuild igrate ome repairs ome is business shelter house -monsoon repairs-winter

Severel hit risis hit it with heav losses it

48 Aid and Recovery in Post-Earthquake Nepal

Borrowing volumes have also grown sub- to the level of earthquake impact. Borrowing is higher stantially since the immediate aftermath of in severely and crisis hit districts (NPR 225,827 and the earthquake. NPR 200,229 on average for each person who borrowed) relative to the lower impact categories (NPR Average sums borrowed have increased from NPR 172,592 and NPR 167,070, respectively). Volumes have 61,440 per person who borrowed in IRM-1 to NPR increased the most in the severely hit districts (402%), 208,749 in IRM-2, a jump of 240% (Table 4.1). As with suggesting that credit is being sought in increasing shares borrowing, loan amounts appear to be related amounts to cope with the impacts of the earthquake.

Table 4.1: Average amount borrowed (NPR) per borrower – by district impact (IRM-1/IRM-2 comparison)

Impact IRM-1 (NPR) IRM-2 (NPR) % Increase Severely hit 44,941 225,827 402% Crisis hit 87,545 200,229 129% Hit with heavy losses 110,959 172,592 56% Hit 34,375 167,070 386% All districts 61,440 208,749 240%

There is, however, large variation in the amounts also in the severely hit category, have borrowed only borrowed between districts within each category of NPR 90,809 on average, the lowest across all districts. impact (Table 4.2). Amongst severely hit districts, for Kathmandu follows Dhading as the district with the example, borrowers in Dhading have taken loans as largest average loan size among borrowers (NPR high as NPR 645,171 on average, the largest amount 528,477), while the other two crisis hit districts have of any district. In contrast, borrowers in Ramechhap, far lower average borrowing amounts.

Table 4.2: Average amount borrowed (NPR) per borrower – by district impact and district (IRM-2)

Average borrowing among Proportion borrowing borrowers (NPR) Severely hit 52% 225,827 Dhading 52% 645,171 Gorkha 47% 159,561 Nuwakot 43% 153,287 Ramechhap 68% 90,809 Sindhupalchowk 51% 111,522 Crisis hit 37% 200,229 Bhaktapur 22% 213,808 Kathmandu 19% 528,477 Okhaldhunga 70% 103,698 Hit with heavy losses 25% 172,592 Lamjung 21% 228,662 Solukhumbu 29% 131,100 Hit 43% 167,070 Syangja 43% 167,070 All districts 42% 208,749

Ramechhap (a severely hit district) and Okhaldhunga mean that while many borrow, they are only eligible (a crisis hit district), which have the highest shares of for smaller loan amounts, leading to lower coping borrowing, have among the lowest average borrowing capacity. amounts. Lower income levels—these two districts have among the largest share of poor households—

49 Coping Strategies

Borrowing is likely to further increase in the insufficient—or if payments were delayed for too long, future. they would have to take on additional debt.

Many households interviewed stated that they would Larger shares of the population in higher impact borrow more money when that began to reconstruct districts intend to borrow in the next three months their homes. If government cash grants prove inade- (Table 4.3). The share is 57% in severely hit districts, quate to pay for housing reconstruction—most people 31% in crisis hit districts, and 21% in the third and felt the NPR 200,000 grant from government would fourth categories.

Table 4.3: Intention to borrow in the next three months – by district impact and district (IRM-2)

Borrow in next three District Current borrowing months Severely hit 52% 57% Dhading 52% 50% Gorkha 47% 50% Nuwakot 43% 49% Ramechhap 68% 79% Sindhupalchowk 51% 58% Crisis hit 37% 31% Bhaktapur 22% 29% Kathmandu 19% 10% Okhaldhunga 70% 54% Hit with heavy losses 25% 21% Lamjung 21% 20% Solukhumbu 29% 21% Hit 43% 21% Syangja 43% 21% All districts 42% 40%

Sources of borrowing have changed since While moneylenders are becoming less prom- IRM-1 with a larger share of people turning to inent in severely hit districts, they continue to formal and semi-formal sources of credit and lend the largest amounts per borrower. a lower share using informal sources. In severely hit districts, they lend on average NPR Relatives and neighbors continue to account for 109,326 to each borrower, almost three times the the largest share of lenders (Figure 4.3). There has amount that banks lend (NPR 35,529 on average). also been a rise in the relative prominence of banks, They are followed by banks, neighbors, relatives, savings groups, and other financial institutions as cooperatives, and savings groups. This ordering is also lenders. This is a positive development given the true overall across the whole sample.33 higher rates of interest charged by informal sources (see Figure 4.4 below). However, it should be noted Interest rates charged on loans have risen that because more people are borrowing than in the slightly for most lenders since IRM-1. IRM-1 period, the actual number of people taking loans from moneylenders has increased: from 3.6% Figure 4.4 shows that the highest rates are charged by of people in IRM-1 to 5.5% in IRM-2. In districts with individuals (on average, 2.51% per month), followed more poor people, moneylenders are more important by moneylenders (2.44%) and neighbors (2.17%). as a borrowing source.32 Interest charged by formal lenders—such as banks, cooperatives, and other financial institutions—are still high relative to what might be expected from

32 See Table 3.6 in IRM-2 survey report. 33 See Figure 3.4 in IRM-2 survey report.

50 Aid and Recovery in Post-Earthquake Nepal

commercial banks.34 Reported bank lending rates have (2.51%) and moneylenders (2.44%). These have also remained consistent since the period immediately risen only slightly since IRM-1 (in the range of 0.1- following the earthquake, with cooperative interest 0.29%). The figures for informal lending institutions, rates rising by 0.1%. Interest rates charged by informal such as moneylenders, appear to be accurate as lending sources in IRM-2 are: friends (1.93%), they are consistent with figures estimated during relativesfigure (2.03%), neighbors (2.17%), individuals qualitative field research.

Figure 4.3: Sources of borrowing among those who borrowed (IRM-1/IRM-2 comparison)

0 29%

23% 2 21% 21%

19% 19%

15 15% 15% 13% 13% 13% 11%

7% 5% 5%

2% 2% 1% 0 0% 0% Friend elative eighbor Individual Bank Savings group oop Finane one- overnment institution lender

IRM-1 I-2

figure 5

Figure 4.4: Monthly interest rates for different sources (IRM-1/IRM-2 comparison)

26 2.51% 2.44% 2.33% 2.25% 2.17% 2.07% 2.03% 2.01% 15 1.93% 1.78% 1.74% 1.74% 1.68%1.71% 1.68% 1.67% 1.61% 1.6%

065

0 Friend elative eighbor Individual Bank Savings group oop Finane onelender institution

IRM-1 I-2

34 The reported rates charged by banks are surprisingly high. Banks of 14-16% for certain types of loans (as per published sources in referred to here are likely to be cooperative banks, which are more 2016). “Cooperatives”, a separate survey option, can refer to formal common in rural areas and tend to have annual interest rates cooperative banks and informal savings groups.

51 Coping Strategies

The disabled were just as likely to borrow but Women were as likely to take loans as men take far smaller loans. but borrow smaller amounts.

For those respondents with a disability, there were There were not notable differences in the proportion no differences in the likelihood of borrowing between of men and of women who borrow (43% of women those with and without disabilities (42% of each have borrowed against 42% of men). However, of those borrowed) and for any particular borrowing purpose. who do borrow, men borrow on average more than However, there is a large difference in borrowing double the amount women do: NPR 288,206 versus amounts: NPR 250,748 for those without disabilities NPR 131,606. Women also reported a lower intention against NPR 143,506 for those with disabilities of to borrow over the next three months (38% against any kind. In addition, the share of those intending to 42% for men). There was very little differences in the borrow in the next three months is much higher for success rates of men and women in securing loans, those with disabilities (45% against 37%). There were and only minimal differences in the sources of loans. no significant differences in sources of borrowing or Borrowing by women from informal and semi-formal rates charged. sources is higher for women: relatives, neighbors, and savings groups (which often are women-specific). Men were more likely to borrow from cooperatives and moneylenders (Figure 4.5). figure 11

Figure 4.5: Sources of borrowing among those who borrowed – by gender (IRM-2)

24% 22% 2 22% 20% 20%

18% 16% 15 15% 15% 15% 14% 12% 13% 12%

7% 6% 6%

0 Friend elative eighbor Individual Bank Savings group oop Finane onelender institution

ale Female

Difference in ethnic or caste identity had some but higher caste people on average take the largest effect on borrowing. loans (NPR 368,249), followed by Janajatis (NPR 117, 534), with lower caste people borrowing much less on Janajatis are less likely to borrow (40%) than low average (NPR 86, 849) – Table 4.4. and high caste people (46% and 45%, respectively),

Table 4.4: Proportion borrowing and amount borrowed – by caste (IRM-2)

Average amount borrowed Caste Proportion borrowing (NPR) Low caste 46% 86,849 Janajati 40% 117,534 High caste 45% 368,249

52 Aid and Recovery in Post-Earthquake Nepal

A much larger share of lower caste people can be found interest rates than Janajatis and higher caste people in the lowest two income brackets (52%) relative to across every type of lender, other than individuals and Janajatis and higher castes (44% and 33%, respective- cooperatives.35 ly). This would explain the lower borrowing amounts among lower castes. Lower caste people also report Higher income households take out much a higher share of past loan refusals (9% compared to larger loans than low or middle income 6% for Janajatis and 5% for high caste), indicating households. credit constraints. However, the lower caste group has a lower share reporting the intention to borrow Those in the top income bracket who borrow take in the next three months (33% versus 40% and 43%). loans of NPR 725,679 on average. This is twice as much as those in the next income bracket and more Lower caste people are the most likely to borrow from than ten times the size of those who had income of less moneylenders (23%) and neighbors (33%); higher than NPR 10,000/month before the earthquake. The caste groups are the most likely to borrow from lowest income group is the most likely to borrow from individuals (18%), cooperatives (16%), and banks moneylenders; and the highest income group is the (15%); and Janajatis are the most likely to borrow from most likely to borrow from banks (36%), individuals relatives (22%), savings groups (21%), and friends (36%), and savings groups (25%) – Table 4.5. (7%). Lower caste people are also charged higher

Table 4.5: Sources of borrowing among those who borrowed – by income band (IRM-2)

Monthly income Friend Relative Neighbor Individual Bank Savings group Cooperative Finance institution Moneylender

NPR 40,000 7% 21% 7% 36% 36% 25% 7% 4% 7% Total 7% 21% 23% 15% 13% 19% 15% 5% 13%

4.2 Remittances

Remittances have increased in importance as affected them (Table 4.6).36 The rate is highest in an income source. Kathmandu (20%) and Bhaktapur (14%) along with a number of severely hit districts (Gorkha, Ramechhap, Remittances are a main source of income for 16% of and Sindhulpalchowk). However, rates of recovery are the population in IRM-2 (against 11% in IRM-1). Nine also highest in the severely hit districts as well as the percent of those who received remittances before the more urban Kathmandu and Bhaktapur. earthquake report that the earthquake negatively

35 See Figure 3.15 in IRM-2 survey report. the ability of people to send remittances. Changes in remittances 36 The World Bank reports that remittances fell in 2015. This was observed here are thus not necessarily linked to the earthquake. primarily a result of a drop in global oil prices, which affected World Bank (2016). Nepal Development Update May 2016: Remittances at Risk. Washington, D.C: World Bank.

53 Coping Strategies

Table 4.6: Remittances as a share of main income source, impact on remittances, and recovery of remittances – by district impact and district (IRM-2)

Share of affected whose Share of people whose Remittance as a main remittances have improved remittances were affected income source (IRM-2) in the past three months by earthquakes (IRM-2) (IRM-2) Severely hit 15% 12% 58% Dhading 24% 8% 86% Gorkha 15% 13% 43% Okhaldhunga 13% 9% 75% Ramechhap 15% 18% 60% Sindhupalchowk 10% 12% 25% Crisis hit 8% 12% 39% Bhaktapur 6% 14% 67% Kathmandu 6% 20% 50% Nuwakot 13% 2% 0% Hit with heavy losses 17% 5% 33% Lamjung 23% 3% 0% Solukhumbu 11% 8% 67% Hit 41% 6% 38% Syangja 41% 6% 38% All districts 16% 9% 53%

While remittances have become more impor- to do so in similar amounts. Thirteen percent say tant as an income source, absolute levels of they continue to receive remittances but at lower remittances from abroad do not appear to volumes than before the earthquake, while 1% say have changed much. that remittances have increased since the earthquake.

Of the 23% who have received remittances from The qualitative research also found little change in abroad, 9% report that these are new remittances levels of remittances. Field interviews found that that began following the earthquake, suggesting that since the monsoon remittances had increased in five the money is intended to cope with disaster impacts wards visited during the research, while in two wards (Figure 4.6). A further 76% who have received remittances decreased. In the other 29 wards there remittancesfigure 1 from abroad report that they received was either no change or no clear trend within the remittances before the earthquake and continue community.

Figure 4.6: Changes in remittances received from abroad (IRM-2)

ont know 1%

eeived remittanes before the earthquake and reeive 1% more sine the earthquake

eeived remittanes before the earthquake and reeive 76% similar level sine the earthquake

eeived remittanes before the earthquake but reeive 13% less sine the earthquake

idnt reeive remittanes before the earthquake 9% but have sine the earthquake 0 20 40 60 0

54 Aid and Recovery in Post-Earthquake Nepal

Findings from the qualitative research suggest that the were found in which family members had returned decrease in remittances from abroad may be because earlier than planned from abroad, or had delayed an some people overseas have returned home to help intended departure to help with reconstruction (see with recovery and reconstruction. Multiple households below).

4.3 Migration

The use of migration as a coping strategy to higher income levels (NPR 10,000 and above). There deal with earthquake impacts has been low. were no significant differences between migration by respondents with disabilities, or for different ethnic Across all districts surveyed, 6% of households mi- or caste groups. grated after the earthquake. This figure includes both internal migrants as well as those who chose to go Of those who migrated since the earthquake, abroad. Eighty-seven percent of those who migrated 60% had returned to their homes by the time did so in the first three months after the earthquake, of the IRM-2 survey. 8% migrated during the 2015 monsoon, and 4% migrated after. Crisis hit districts, which include Kath- This finding was substantiated by the qualitative mandu and Bhaktapur, have more than double the research. Many people who were initially displaced— rate of migration on average (12%) relative to severely as they sought areas with less landside risk, or hit districts (5%); the third and fourth categories of temporary shelter for the monsoon—were returning impact have lower rates (4% and 1%, respectively). to their communities, often to cultivate their fields.

The most commonly cited reason for migration was In terms of international migration, the qualitative lack of shelter (68%), followed by lack of livelihood research found stories of people who were working opportunities (22%), landslides caused by the abroad returning early and those who were planning earthquake (17%), and the risk of future landslides to leave delaying their departure to support the recon- (10%) – Figure 4.7. struction efforts. There were isolated instances and the field research did not clearly establish a wider trend of migrants returning home in the months since the figureFigure 20 4.7: Reasons for migration (IRM-2) earthquake. However, such stories are corroborated by official statistics.37 Rates of international migration 0 68% were already high in some areas studied, and there was no clear indication that more households than usual are considering migrating abroad.

22%

1 17%

10% 6% 1% 0% 0 arthquakeandslide o shelter o livelihoodain eath isease

37 Official statistics suggest a trend of migrant workers returning to take care of their families. See Ratha et.al. 2016 “Migration and Remittances – Recent Developments and Outlook” Migration The percentage of households reporting migration was and Development Brief 26, April 2016, World Bank, Washington, lower in rural areas compared with urban areas (5% DC http://pubdocs.worldbank.org/en/661301460400427908/ against 9%). There was also higher migration rates at MigrationandDevelopmentBrief26.pdf

55 Coping Strategies

4.4 Asset Sales

Sale of assets since the earthquake has been across the whole sample have sold livestock. Of those low. who sold livestock, 17% say they sold all, 28% over half of the livestock they had, 30% sold between one-quar- Across all districts, 6% of those surveyed had sold ter and one-half of their livestock, and 26% under assets since the beginning of the 2015 monsoon. This one-quarter. Qualitative interviews found examples was higher in severely hit districts (average 8%), of families selling livestock due to displacement as especially Ramechhap and Sindhupalchowk, and also families in high impact wards moved to temporary in Okhaldhunga and Solukhumbu (Figure 4.8). settlements. Other survivors reported selling their livestock to meet immediate needs such as household The vast majority of asset sales (89%) have expenses and to pay for building temporary shelters been of livestock. or to repair damaged structures. Some respondents described the slowing of asset sales after the monsoon Given that 6% of all people sold assets, and 89% of because of the cash grants provided by the govern- these sold livestock, this means that 5.3% of all people ment, which helped people address their basic needs. figure 22

Figure 4.8: Share of people selling assets – by district (IRM-2)

16 16%

14%

12 11% 10%

4 4% 3% 2% 2% 1% 1% 0 hading orkha uwakot amehhap SindhupalhowkBhaktapur athmandu khaldhunga amung Solukhumbu Sanga

Land sales have been relatively low. waiting for government grants before deciding to take such drastic measures, and also that in severely People selling land account for 8% of those selling affected areas there was not a significant market of assets. This means that around 0.5% of all respondents people trying to purchase land. Qualitative interviews have sold land. Of this group, 10% had sold all their suggest, however, that many households will consider land, with 3% selling over half, 28% selling between selling land in the future if they cannot repay their one-quarter and one-half, and 59% less than one- loans, or if government grants are not sufficient to quarter. Qualitative field research suggests that land rebuild their homes. sales have been low because many households were

56 Aid and Recovery in Post-Earthquake Nepal

Chapter 5. Politics and Leadership

Photo: Tanka Gurung

In the wake of the earthquake, local officials, political party representatives in decision-making. While party leaders and the national government all played Ward Citizen Forums have played a more active role, critical roles in responding to emergency needs. IRM-2 there has not been a rise in new leaders. Satisfaction found that coordination and leadership of the recovery with political parties has declined but there is no effort at the local level now reflects common practice clear indication yet of changes in people’s political in Nepal, with local officials consulting political preferences.

5.1 Political parties

Political parties continue to play a major role Political parties were sufficiently organized to listen in decision-making around aid. to local needs through their representatives at the ward and VDC levels, and to then communicate that Local officials at the ward, VDC, and district levels information to stakeholders at the district headquar- have consulted political parties in local decision mak- ters through their participation in DDRCs and other ing for many years, and this has continued since the meetings. Similarly, these communication channels earthquake. Political party leadership were formally within political parties were at times useful for inform- included in the DDRCs and RDCs after the earthquake. ing local citizens about decisions made at the district During the emergency relief period, they played an im- headquarters and in Kathmandu, as well as informing portant role assisting overburdened local bureaucrats NGOs on aid needs and providing information for in planning and executing the government’s immedi- targeting. ate relief and recovery programs, and in mediating earthquake-affected people’s grievances with the state. However, dissatisfaction with the role politi- During the second round of field research, political cal parties played in responding to post-earth- parties were still involved in DDRCs and RDCs, al- quake needs was high and has increased in the though these bodies were less active compared to the second round of research. pre-monsoon period last year. Overall, satisfaction with the role of local political Political parties facilitated communication parties in the aid response is low. Only 32% of re- between local residents and government of- spondents in IRM-2 reported feeling satisfied with ficials at the district level. parties with 61% reporting dissatisfaction. Those living

57 Politics and Leadership

in more affected districts were even more likely to local political parties. Of those who were very satisfied express dissatisfaction with political parties than those in IRM-1, 62% now state dissatisfaction; 57% of those in areas that were impacted less (Figure 5.1). who said they were somewhat satisfied in IRM-1 now express dissatisfaction. Only 30% or less of respond- Comparing responses between IRM-1 and IRM-2, ents who expressed dissatisfaction in the first phase there is a clear trend of decreasing satisfaction with of research have become more satisfied (Table 5.1).

Figure 5.1: Satisfaction with local political parties – by district impact (IRM-2)

8% 49% 14% 24% 5% it

16% 24% 25% 28% 8% it with heav losses

2% 17% 26% 47% 8% risis hit

3% 27% 25% 38% 7% Severel hit

0 25 50 5 100

er satisfied er unsatisfied Somewhat satisfied ont know Somewhat unsatisfied

Table 5.1: Satisfaction with local political parties – individual panel data (IRM-1/IRM-2 comparison)

How satisfied or unsatisfied are you with local political parties? (IRM-2) Total Very Somewhat Somewhat Very Don’t satisfied satisfied unsatisfied unsatisfied know Very satisfied 11% 24% 38% 24% 3% 100% How satisfied or Somewhat satisfied 6% 32% 23% 34% 4% 100% unsatisfied are Somewhat unsatisfied 3% 23% 30% 39% 5% 100% you with local Very unsatisfied 4% 26% 22% 41% 6% 100% political parties? Refused 17% 17% 17% 50% 100% (IRM-1) Don’t know 2% 24% 20% 37% 16% 100%

A key reason for dissatisfaction is perceptions Declining levels of aid have led people to that damage assessments were politicized. blame political parties.

In IRM-2 complaints about political actors influencing The level of aid provided in a location seems to be the outcome of damage assessments were more an important determinant of levels of satisfaction frequent, and people commonly felt that inflated or with local political parties. This can be seen in the understated damage assessments were the result of variation between districts within impact categories in political party favoritism. satisfaction with political parties’ role in aid delivery (Table 5.2). Amongst the severely hit districts, there is less dissatisfaction in Ramechhap and Nuwakot than in other districts. Ramechhap and Nuwakot

58 Aid and Recovery in Post-Earthquake Nepal

have received more aid than most other severely hit in Lamjung. The exception is Sindhupalchowk, which districts. For hit with heavy losses districts, people received high amounts of aid but where satisfaction are much more likely to be satisfied in Solukhumbu, with political parties remains low. which has received significant amounts of aid, than

Table 5.2: Satisfaction with local political parties – by district (IRM-2)

How satisfied or unsatisfied you are with local political parties? Somewhat Somewhat Very Very satisfied Don’t know Total satisfied unsatisfied unsatisfied Severely hit 3% 27% 25% 38% 7% 100% Dhading 1% 20% 11% 66% 2% 100% Gorkha* 3% 21% 23% 49% 4% 100% Nuwakot 3% 26% 31% 25% 15% 100% Ramechhap 4% 39% 33% 18% 5% 100% Sindhupalchowk 5% 28% 25% 36% 7% 100% Crisis hit 2% 17% 26% 47% 8% 100% Bhaktapur 1% 17% 14% 57% 10% 100% Kathmandu* 2% 16% 23% 53% 5% 100% Okhaldhunga 2% 17% 43% 29% 9% 100% Hit with heavy losses 16% 24% 25% 28% 8% 100% Solukhumbu 30% 21% 15% 23% 11% 100% Lamjung 1% 26% 34% 33% 5% 100% Hit 8% 49% 14% 24% 5% 100% Syangja 8% 49% 14% 24% 5% 100% *1% refused to respond

While satisfaction with political parties has government in IRM-1 expressed satisfaction with been declining, this is not true for overall the central government in IRM-2, whereas 32% of perceptions of government performance. respondents who were satisfied in IRM-1 are now somewhat or very dissatisfied. This net positive trend Satisfaction with the central government increased took place despite the fact that 74% of households between IRM-1 and IRM-2. Using the panel data, reported that their constituent assembly members had 55% or more of respondents who reported they were not visited since the 2015 monsoon season.38 either somewhat or very unsatisfied with the central

Table 5.3: Satisfaction with the central government – individual panel data (IRM-1/IRM-2 comparison)

How satisfied or unsatisfied are you with central government? (IRM-2) Total Very Somewhat Somewhat Very Don’t know satisfied satisfied unsatisfied unsatisfied

How satisfied Very satisfied 14% 53% 16% 16% 1% 100% or unsatis- Somewhat satisfied 15% 50% 17% 14% 3% 100% fied are you Somewhat unsatisfied 13% 45% 19% 19% 3% 100% with central Very unsatisfied 7% 48% 21% 22% 2% 100% government? Refused 0% 100% 0% 0% 0% 100% (IRM-2) Don’t know 16% 49% 12% 16% 8% 100%

38 See Figure 7.13 in IRM-2 survey report.

59 Politics and Leadership

5.2 Emergence of new leadership and political party preferences

There is no evidence of new leadership emerg- There are indications of some changes in ing that is challenging existing political political preferences but most people remain dynamics. undecided as to who they will vote for in the next election. In many places, Ward Citizen Forum (WCF) coordinators have played an increasing role since IRM-1. In Among those who have decided who they will vote for, one-third of VDCs visited for the qualitative fieldwork, there is increased support for Nepali Congress and researchers observed that the WCF had been an im- CPN-UML with a slight drop in support for UCPN portant local institution during relief and recovery (Maoist). Table 5.4 combines information on who and that WCF coordinators had played leadership people voted for before and who they say they will roles during the implementation of recovery efforts. vote for at the next election. As with IRM-1, of those In wards where the presence of political parties was who have decided who to support in the next election, rather limited, such as in parts of Gorkha district, WCF most intend to vote for the same party as before. coordinators were more active in the planning and For example, only 1% of people who voted Nepali distribution of relief. However, there was no indication Congress in the last election say they will vote for a that this is spilling over into an enhanced role for WCF different party next time round. The figure is also 1% coordinators in broader local governance. WCF mem- for CPN-UML. Past UCPN (Maoist) supporters make bers were quick to point out to researchers that they up a larger share of voters in severely hit districts. But remain marginal in actual decision-making. Informal only 20% of those who said they chose UCPN (Maoist) local leaders, including general citizens, teachers, sav- in the last election state that they would vote for the ing and credit cooperative officials, philanthropists, party if an election were held. Indeed, those who voted and business people, have become involved in the for UCPN (Maoist) in the last election are more likely relief and recovery process. But this has not translated to have shifted their allegiance, with 8% saying they into wider leadership roles. will now vote for Nepal Congress or CPN-UML and 61% saying they still have to make up their mind. The decline in support for UCPN (Maoist) is seen across all caste groups.

Table 5.4: Current political preferences – by past votes (IRM-2)

If an election was to be held soon, which party would you vote for? (IRM-2) Nepali Congress CPN-UML UCPN (M) RPP-N RPP MJF-Nepal Will not vote NMKP Refused know Don’t Total

Nepal Congress 49% 1% 0% 0% 0% 0% 3% 0% 4% 42% 100% CPN-UML 1% 44% 0% 0% 0% 0% 4% 0% 6% 45% 100% UCPN (Maoist) 5% 3% 20% 0% 0% 0% 4% 0% 6% 61% 100% Voted for RPP-N 0% 8% 0% 52% 0% 0% 0% 0% 4% 36% 100% in last RPP 5% 3% 0% 0% 40% 0% 3% 0% 3% 48% 100% election MJF-Nepal 0% 0% 0% 0% 0% 75% 25% 0% 0% 0% 100% (IRM-2) Did not vote 5% 3% 1% 1% 0% 0% 18% 0% 4% 68% 100% NMKP 0% 0% 0% 0% 0% 0% 6% 26% 4% 64% 100% Refused 0% 1% 0% 0% 0% 0% 2% 0% 53% 44% 100% Don’t know 0% 0% 0% 0% 0% 0% 1% 0% 1% 98% 100%

60 Aid and Recovery in Post-Earthquake Nepal

Chapter 6. Social Relations and Conflict

Photo: Amanda Gurung

Continuing strong social cohesion characterizes a small number of communities, there has been almost earthquake-affected areas. Perceptions of safety have no violence. Tensions are largely related to perceived increased since IRM-1. Despite increasing tensions in unfairness in aid distribution and displacement.

6.1 Security and crime

Feelings of safety have improved since IRM-1. Perceptions of safety did not vary for men and women, or by other factors, except for where Whereas 83% of respondents in IRM-1 reported feeling people are living now. safe in their communities, this has increased to 97% of IRM-2 respondents. Qualitative field interviews with The survey results do not show differences in per- security officers and community members suggest that ceptions of safety by gender. Likewise, there are no there has not been any significant change in crime significant variations when disaggregating perceptions rates during the monsoon or the winter season.39 of safety by caste or disability. Households in urban areas felt very slightly less secure (96% safe in urban People in severely hit districts report feeling safer com- areas and 98% in rural locations). The most significant pared to other areas, but differences between areas with variable for perceptions of safety was the type of shel- different earthquake impacts is minimal (Figure 6.1). ter households people lived in. Those living away from Increased feelings of safety have occurred despite the their own land felt more unsafe. This included those deployment of security personnel normalizing, having in temporary shelter on other people’s land, people increased in affected areas after the disaster. Only renting houses, and those living in temporary shelters Sindhupalchowk district had yet to see a decrease in on public land (Table 6.1). Qualitative interviews also security personnel back to pre-earthquake levels. found heightened perceptions of insecurity among

39 Some security officers said they believed some criminal activities There has been a slight increase in official reports of rapes and that had been an issue prior to the earthquake, such as robberies suicides in some districts, but local security actors do not believe and illegal extraction from community forests in Gorkha, and petty there has been a systematic increase. crimes in Sindhupalchowk, had decreased since the earthquake.

61 Social Relations and Conflict

those in temporary shelter, most notably women. of being exposed to the risk of sexual violence and Some women living in temporary shelter who were robbery. interviewed during the field research expressed fear

Figure 6.1: How safe and secure do you feel now in your community? – by district impact (IRM-2)

48% 49% 2% it

49% 50% 1% it with heav losses

56% 41% 3% risis hit

64% 34% 2% Severel hit

0 25 50 5 100

er safe Somewhat safe Somewhat unsafe

Table 6.1: How safe and secure do you feel now in your community? – by current shelter (IRM-2)

Somewhat Somewhat Very Where are you living now? Very safe Total safe unsafe unsafe Own house 53% 44% 2% 0% 100% Neighbor’s house 44% 53% 0% 3% 100% Friend’s house 67% 33% 0% 0% 100% Self-constructed shelter on own land 64% 34% 2% 0% 100% Self-constructed shelter on other people’s land 68% 28% 4% 0% 100% Self-constructed shelter on public land 58% 33% 8% 0% 100% Community shelter 67% 33% 0% 0% 100% Rent 60% 27% 7% 7% 100%

There were almost no reports of violence While official statistics and quantitative sur- occurring. vey results suggest violence and crime has reduced, issues related to gender-based vio- Ninety-nine percent of people report that there have lence and alcohol intake should be monitored. been no violent incidents in their community since the beginning of the 2015 monsoon. Only in Syangja Many respondents reported that they believed inci- district did more than 1% of households report that dents of sexual and gender-based violence, including a violent incident had occurred in their community domestic violence, had occurred since the disaster. since the monsoon.40 In IRM-1, 3% of households These crimes were clearly taking place before the reported that violent incidents had occurred since earthquake as well and interviews could not establish a the earthquake. clear earthquake-related trend. Respondents in Sind- hupalchowk, Ramechhap, Okhaldhunga, Solukhum- bu, and Syangja districts all reported perceptions that over the course of the monsoon and winter seasons alcohol consumption had increased, in part as a coping 40 The figure was 1.4% in Syangja. mechanism to deal with trauma and poor living condi-

62 Aid and Recovery in Post-Earthquake Nepal

tions. Security officials also frequently cited alcoholism corroborate, they present potential emerging challeng- as a cause of violence and crimes, especially domestic es that will need to be tracked.41 violence. While these perceived trends are difficult to

6.2 Trust and social cohesion

In most communities, social relations have despite being under multiple pressures, and this trend remained strong. has continued since the emergency phase ended. There have been many complaints, such as those over Out of 36 wards visited in the qualitative study, 27 damage assessments, but communities have managed wards reported that social cohesion did not deteriorate tensions without them escalating into broader social after the earthquake (Figure 6.2). Most communities problems. maintained good social relations after the earthquake,

figure 51 Figure 6.2: Assessment of trends in social relations post-earthquake by ward respondents

6 6 6 6 6 6 5

3 3

0 0 0 Severel hit risis hit it with heav losses it

Wards with stableimproved soial relations Wards with reports of soial tensions Total number of wards in eah ategor

In seven of these communities, social cohesion has to help each other. For example, Gurung families improved since the earthquake. People attributed in one ward in Syangja allowed six Dalit families this to a spirit of cooperation and communal living to live in unoccupied houses through the monsoon as they coped with challenges in the immediate season. However, most of these cases occurred in aftermath of the earthquake. In one ward studied the early months after the disaster, and it is unclear in Sindhulpalchowk, for example, 85 families whether there will be lasting positive impacts on social lived together in a communal shelter following the relations. earthquake. They shared a common meal, sharing 60kg of mixed porridge among themselves equally, and developed a system of rotating labor exchanges to build temporary shelters, clear rubble, and resume 41 IRM-2 did not reveal incidents of trafficking during the field farming. In another wards, people provided low research, although other studies have noted that trafficking may interest loans to affected neighbors. There were cases have increased after the earthquake in certain areas. The situation of people going across caste and ethnic boundaries should be monitored carefully.

63 Social Relations and Conflict

In nine other communities studied, there are cases had a perceived caste or ethnic dimension to signs of significant social tensions but this has them, with discrimination against Dalits, low caste not led to open conflict or violence. community members, or involving tension between ethnic groups. While these tensions were present, The cases involved three issues: (i) displacement and conflict and disagreements remained limited to verbal resettlement; (ii) perceived discrimination or unfair confrontation or complaints and resentment. In no treatment in relief distribution; and (iii) conflict over cases did field researchers find evidence of escalation scarce water.42 Observed conflicts and tensions, with and violence. Disputes around displacement and the exception of one case related to water resources, resettlement, especially when it related to the generally involved friction between a minority (ethnic, occupation of public land, were seen as causing the caste, displaced persons) and a majority group. Several most tension.

Case Study 6: Tensions over relocation in Sinhulpalchowk Citizens relocating without adequate consul- exploit the economic opportunity of occupying tation with local officials or with the wider land on the main road. community created tension in some locations. In Syaule Ward 8 in Sindhupalchowk, some Since Kerabari residents had also experienced tension was generated when 19 families moved deaths during the earthquake due to landslides, from Dadagaun because of fears of landslides. and had also suffered from landslide problems They set up shelters on the main road into the before the earthquake, their frustration was VDC, occupying potentially valuable commer- heightened. Those interviewed explained that cial land and naming the location Naya Basti. households in their community had already filed Although they reached out to the RDC to seek a claim for resettlement before the earthquake. permission from the DDC for the resettlement, They strongly felt that they had suffered more and were still awaiting an answer at the time of damage than the Naya Basti residents who were the research, other VDC residents thought they poised to benefit economically by establishing were being opportunistic and intentionally us- a new settlement on the main road of the VDC. ing the earthquake as an excuse to permanently Given these tensions, people in Naya Basti occupy the land. Giving their settlement a name kept hearing rumors in the village that they only strengthened the suspicion that they were would have to move soon and that the owners planning to settle there permanently. Other of the land on which they had settled planned villagers, including those from a nearby settle- to sell the land. As an official decision on their ment called Kerabari, resented this, accusing settlement was pending, the relations between the people from Dadagaun of relocating with- the Naya Basti residents and the rest of the out having been at real risk, and of planning to village were very poor.

42 More details on these and other cases can be found in Chapter 8 of the IRM-2 qualitative report.

64 Aid and Recovery in Post-Earthquake Nepal

6.3 Sense of fairness and grievances with aid providers

It is widely held that people of all identities when disaggregated by impact levels, there is more have had equal access to aid. substantial variation (Figure 6.3). Ten percent of respondents in hit and crisis hit districts say they Overall, 89% of surveyed respondents agreed that disagree, compared to 6% in hit with heavy losses people of every caste, religion, and ethnicity have districts and severely hit districts. There is relatively been equally able to receive aid according to their little difference in the responses of people in urban needs. Across all areas, only 8% disagreed. However, and rural areas, or of men and women.

Figure 6.3: Do you think people of every caste, religion, and ethnicity are equally able to receive aid according to their needs? – by district impact, urban/rural, and gender (IRM-2)

20% 73% 6% 1% Severel hit

21% 60% 9% 2% 8% risis hit 10% 83% 5% 2% it with heav losses 25% 65% 10% it 17% 73% 7% 2% ural area 27% 55% 7% 2% 9% rban area 17% 72% 7% 4% Female 20% 70% 6% 1% 3% ale

0 25 50 5 100

Strongl agree Strongl disagree Somewhat agree ont know Somewhat disagree

More people think that aid is distributed fairly hit districts were the most likely to feel aid is being by VDCs /municipalities than in the early distributed fairly while those in the crisis hit districts months after the earthquake. Perceptions of were the least likely (Table 6.2). This is likely because fair distribution are more common in severely aid distribution has been very wide in severely hit hit districts. districts (reaching 95% of people) whereas only around one-half of people in crisis hit districts have received There is a large increase in the share of people who aid. This link between limited distribution and feel that aid is being distributed fairly by VDCs/ perceived unfairness can be seen in Lamjung district, municipalities. In IRM-1, 55% of people felt aid was where 45% reported that aid was not distributed being fairly distributed.43 When asked if this was fairly.44 Aid distribution has been particularly low in the case since the beginning of the 2015 monsoon, Lamjung, with 53% saying they had not received aid. 76% agreed and 28% disagreed. Those in severely

43 IRM-1 survey report, p. 44. 44 Lamjung is the district that borders with Gorkha, the epicenter of the first earthquake.

65 Social Relations and Conflict

Photo: Alok Pokharel

Table 6.2: Fair distribution by VDC/municipalities – by district impact and district (IRM-2)

Strongly Somewhat Somewhat Strongly Refused Don’t know Total agree agree disagree disagree Severely hit 34% 42% 12% 9% 0% 3% 100% Dhading 34% 38% 13% 10% 0% 5% 100% Gorkha 33% 43% 9% 14% 0% 2% 100% Nuwakot 35% 45% 11% 6% 0% 2% 100% Ramechhap 27% 49% 15% 6% 0% 3% 100% Sindhupalchowk 39% 35% 11% 10% 0% 5% 100% Crisis hit 14% 35% 14% 25% 0% 12% 100% Bhaktapur 12% 28% 13% 21% 1% 25% 100% Kathmandu 4% 47% 16% 28% 1% 5% 100% Okhaldhunga 27% 30% 13% 26% 0% 5% 100% Hit with heavy losses 30% 38% 13% 15% 0% 4% 100% Lamjung 10% 40% 20% 25% 0% 5% 100% Solukhumbu 51% 37% 5% 4% 0% 3% 100% Hit 23% 41% 12% 13% 0% 11% 100% Syangja 23% 41% 12% 13% 0% 11% 100%

There were perceptions of unfair distribution those reported during the first round of research. based on location, caste, and perceived politi- There were examples of remotely located communities cal bias in some places. However, frustrations feeling that their access to aid was less than that of were directed at decision-makers and did not more easily accessible wards were present.45 Dalits also lead to violence. reported feeling disadvantaged in aid distribution in at least two wards visited.46 The quantitative survey also Complaints about access to relief, targeting, damage confirmed the higher perception of unfairness among assessments, and other issues were consistent with Dalits (Figure 6.4).47

45 This was heard in wards including Barpak wards 1, 2, and 3 in one listened to their opinions and that they had been discriminated Gorkha, Baruneshwor Ward 1 and Katunje Ward 4 in Okhaldhunga, against in relief and the damage assessments. and Syaule Ward 1 in Sindhupalchowk 47 However, Dalits were more likely to agree that aid was distributed 46 In Dhuwakot Ward 9, Gorkha, Dalits felt that Brahmins and other fairly by VDC/municipalities since the last monsoon (68%), in residents had better access to relief and the best materials because contrast to Madhesi ethnic (36%), Madhesi caste (56%), Newars the relief was dropped off in an area near high caste family houses. (58%), and hill castes (60%). Dalits in Katunje VDC Ward 1 in Okhaldhunga reported that no

66 Aid and Recovery in Post-Earthquake Nepal

Figure 6.4: Do you think people of every caste, religion, and ethnicity are equally able to receive aid according to their needs? – by ethnic group (IRM-2)

6% 94% adhesi ethni

11% 78% 11% adhesi aste

24% 59% 8% 1% 8% ewar

14% 64% 21% 1% ill dalit

18% 75% 4% 1% 2% ill ethni

18% 72% 6% 1% 3% ill aste

0 25 50 5 100

Strongl agree Strongl disagree Somewhat agree ont know Somewhat disagree

Equal distribution of aid was still commonly to members of ‘marginalized’ or ‘vulnerable’ groups, used as it helped maintain good social rela- including Janajati, Dalits, women, the elderly, and tion. Targeted aid was welcome in some plac- children. In some VDCs, dominant social groups ar- es, but it created tensions in other locations. ticulated resentment toward neighbors from different social or ethnic groups who were the target group for The most common pattern of aid delivery from the a specific program’s aid. This happened in Tanglichok RDC was even and equal distribution to all qualifying VDC Ward 9 in Gorkha district, a ward that is pre- households even when they had endured different dominately inhabited by economically disadvantaged levels of impact. In most places, people were happy Chepangs, within a VDC largely inhabited by Gurungs. with this. There were cases where uneven distribution Similarly, the feeling that low caste Dalits are prior- led to disputes that were only resolved when the itized for aid over high caste Brahmin and Chettri decision was taken to return to equal distribution.48 residents was voiced by higher castes in Katunje and Baruneshwor VDCs in Okhaldhunga district. Aid delivered by NGOs and INGOs was more likely to be targeted and was usually distributed directly

6.4. Potential sources of conflict

Scarce water sources affect farming and life, there are also signs that it has the potential to water consumption and can cause conflicts. cause conflict between residents. In Lisankhu Ward 3, Sindhupalchowk, frequent disputes between women Disruption of water sources, damage to irrigation collecting water for their household were reported. systems, and shortages of drinking water were re- Delayed attention to the water needs of affected com- ported in multiple areas in four out of the six districts munities has the potential to lead to more conflict in studied. While this clearly affects livelihoods and daily the future.

48 This happened in both Nele Ward 7, Solukhumbu, and Doramba Ward 5 in Ramechhap.

67 Social Relations and Conflict

Ongoing displacement is a potential source of Patterns of relief distribution can exacerbate conflict in communities where resettlement existing tensions or turn existing social was tolerated because it was to be temporary. prejudices into tensions.

The unplanned nature of resettlement during the 2015 So far, perceived or actual discrimination/unfairness monsoon, particularly concerning the use of public in access to aid has not led to a serious deterioration land, has produced friction in some communities, in social relations in communities. However, if which was expressed variously from resentment resentment is left unaddressed, and especially in areas through to outright caste discrimination. Even in that already experience divisions along communal places where the use of public land has yet to cause lines, there is potential for disputes to escalate into conflict, accommodating displaced persons on both larger intra-community conflict. In a context going private and public land was regarded as a temporary forward where programs target only those who hold a solution for the previous monsoon. As residents of beneficiary card, the challenge of perceptions created Barpak in Gorkha district related, they were happy to by using different targeting strategies will become let others live on their land rent free initially, but they increasingly important to monitor and engage with. would be unable to do so indefinitely.

Photo: Chiran Manandhar

68 Aid and Recovery in Post-Earthquake Nepal

Chapter 7. Conclusions and Recommendations

Photo: Chiran Manandhar

7.1 Overview of conclusions

The first round of IRM, conducted in June 2015, The vast majority of those whose houses were badly or found that the physical impacts of the earthquake had completely damaged are still living in self-constructed been immense. In many areas, almost all houses were temporary shelters. Another difficult monsoon season fully or partially destroyed and public facilities often in 2016 was being anticipated at the time of the field suffered significant damage. Affected communities research. Despite these challenges, the qualitative re- were using a variety of strategies to cope and recover. search came across communities who banded together Emergency aid was helping many, including those in to take care of the worst affected amongst them and remote areas, but there was evidence of mistargeting. of households showing great resilience in the face of At the time of IRM-1, people said they needed new significant challenges. forms of assistance for the coming months, especially cash and support for building sturdier shelters or to The economic strain on households is clear reconstruct their houses. with borrowing increasing rapidly.

IRM-2, conducted almost one year on from the Nepal Borrowing has risen significantly since IRM-1, both in earthquakes, provides evidence of how conditions the number of people borrowing and the size of loans. and needs have evolved since the emergency period. Informal sources for borrowing predominate, espe- A large quantitative dataset provides a clear sense of cially in rural areas, but there has been a partial move trends related to aid and recovery, while robust quali- towards borrowing from formal sources who charge tative data allow for a more in-depth understanding of lower rates of interest. Borrowing is particularly high how different households and communities are coping in poorer districts, and there is some evidence that with earthquake impacts and recovering. The data certain vulnerable groups—such as lower caste peo- show that the slow pace of reconstruction and recovery ple—are often having trouble accessing larger loans. efforts is taking a toll on households and communities, Increased borrowing is occurring despite greater especially those still in temporary shelter. volumes of cash reaching the earthquake-affected, largely through government cash grant programs for Earthquake survivors endured difficult the affected. The amounts received have been insuf- monsoon and winter seasons, often with in- ficient to meet the expenses of households, or to kick adequate shelter, leading to discomfort and start the languishing rebuilding process. By March sickness. 2016, the demand for cash as a priority need had risen

69 Social Relations and Conflict

exponentially along with the need for reconstruction Food aid appears to have been well targeted, reaching materials to help get people back into secure homes. most in severely food insecure areas and at larger volumes than elsewhere. Yet food distributed has been Livelihoods have rebounded, with most insufficient to prevent households from decreasing people back at work. their food consumption. This is an issue to watch as trends from IRM-1 to IRM-2 show reductions in the Most businesses, in particular, have made progress in coverage of food aid. recovery since IRM-1. Farmers, most of the population in affected areas, have gone back to their fields, but While the strain of this slow recovery is start- have seen slower recovery in severely hit districts than ing to emerge in community dynamics and have businesses or daily wage workers. Remittances, public sentiment, social cohesion on the whole migration, and asset sales have been less commonly has not significantly worsened to this point – used coping strategies. but there are risks for the future.

Aid has been critical to households and satis- Psychological impacts from the traumatic experiences faction with what has been provided is gen- of the last year are impacting some families. There are erally high, but IRM-2 shows that aid is still cases of overt discrimination and disputes along ethnic sorely needed. or caste lines, but these appear isolated at present. Coordination and communication challenges underlie Volumes of aid have declined since the early post-earth- much of the negative perceptions of the recovery effort. quake months. The data shows that aid now primarily These perceptions are translating into decreasing comes in the form of cash, at low volumes. Some dissatisfaction with the role of political parties since tricts, in particular Okhaldhunga, have seen especially IRM-1, but overall satisfaction with government has large declines in aid despite having high levels of need not yet shown significant deterioration. for recovery assistance. Okhaldhunga was impacted as much as Solukhumbu, but Solukhumbu has received Targeting aid to specific groups can be contentious if much more aid since IRM-1. Sindhupalchowk, which not managed carefully and tensions have risen since experienced the greatest physical impacts from the IRM-1. This in part reflects increasing challenges in earthquake, has had by far the highest levels of aid – targeting aid as overall aid volumes decrease and the but this has still not been sufficient to meet key needs equal distribution of aid across affected communities such as shelter or food. Aid has also been slow to becomes less viable. In addition, the damage assessment shift from the emergency relief phase to supporting process carried out by the government (in multiple recovery and reconstruction. Emergency types of aid, stages, with varying and opaque methodologies) has such as tarps, were still being provided but there were caused disgruntlement. As the cash grants provided only small amounts of materials for reconstruction increase as reconstruction assistance begins in the being distributed, suggesting an emerging mismatch coming months, the contentious nature of these between what people receive and what they need. assessments has the potential to lead to increased dissatisfaction and diminished community cohesion.

7.2 Implications and recommendations

The data and analysis from the IRM-2 field research ❶ Short and medium term improvements in has established emerging challenges relevant to temporary shelter are needed. ongoing and future assistance for earthquake recovery. The National Reconstruction Authority (NRA) has The extremely high proportion of the population already begun to respond to some issues raised by the still living in temporary shelter in districts that research since starting its work but challenges and were severely hit by the quakes is of concern. risks remain. The paper concludes by providing a set Delays in reconstruction support, and inadequate of independent recommendations for aid providers.49 communication on emerging policies, combined with limited delivery of reconstruction materials, has meant that very little rebuilding has started. The beginning of the formal reconstruction period, which will see the disbursement of larger sums 49 These are independent recommendations rather than those of of cash, will help some people begin the task of the UK or Swiss governments. rebuilding. Yet it is unrealistic to expect that the

70 Aid and Recovery in Post-Earthquake Nepal

majority of people currently in temporary shelters Recommendation 3 – Expand soft loan or partially damaged houses will be able to rebuild programs, strengthen communication in the near future. Limited labor availability, recon- about them, and ensure they reach those struction grants being less than the amount needed in rural areas; to build a new house, and the fact that many people whose houses were partially damaged will not re- Recommendation 4 – Consider options ceive large-scale government funds all indicate that for regulation of interest rates used by housing reconstruction will be a long-term process. informal lenders. There is thus a need to develop short-to-medium term strategies to improve the living conditions of people who will likely remain in shelters for the ❸ Aid providers need to adjust strategies for foreseeable future, and also to respond to the neg- the next phase of earthquake recovery in ative impacts on mental and physical health and order to match aid with evolving needs. the perceived insecurity that those in inadequate shelter suffer from. It is important to ensure that aid planning takes into account emerging needs and that it evolves Recommendation 1 – Develop a strategy along with the recovery process. There is evidence to provide new, or improve the quality of that some aid has been supply-driven with things existing, temporary housing for the me- delivered that are not priority needs for affected dium-term; people. IRM-2 data showed an overwhelming need among the earthquake-affected for cash, housing, Recommendation 2 – Prioritize programs and food. People prefer the former because it is to mitigate the consequences of staying in liquid – people can choose to spend money in ways temporary shelter (targeted health sup- that fit with the distinctive needs they have. Despite port and medicine, temporary water and the need for robust housing, little assistance has sanitation facilities, women’s security). been provided in this area to date. Targeting of food aid has generally been strong, but coverage of food aid is declining and this may cause future ❷ Avoiding debt traps among the earthquake problems. Those needing livelihood support are affected will require more accessible largely in the agriculture sector. opportunities for affordable credit or cash grants. Recommendation 5 – Focus on cash support and housing materials; Borrowing is the most-used coping strategy. However, with livelihoods having not fully Recommendation 6 – Extend food aid to recovered, and capital needed for rebuilding, it ensure food insecurity does not increase; is likely that many will face difficulties in paying back the loans they take. The high levels of Recommendation 7 – Develop strategies interest rates charged by many of the dominant to help farmers recover; providers of loans increases this risk. The NRA is developing a plan to provide housing loans Recommendation 8 – Strengthen commu- and there are already soft loan programs offered nication channels for local communities to by the government. However, access for most express their needs to aid providers who earthquake-affected, given their locations in rural are active in their areas. areas, and challenges for potential recipients in proving credit worthiness, mean that the current programs are not being widely utilized. There ❹ Enhancing communication with affected has also been poor communication regarding communities and local officials is critical. the available programs. Without clear policy and available credit, the potential for an increased Confusion and uncertainty has been one of the reliance on moneylenders, who charge high interest major constraints to recovery. While government rates, is real. There is a need to ensure that credit and donor inputs are important for the recovery is provided at reasonable interest rates, to provide process, it is ultimately the actions taken by additional cash to limit borrowing, and to provide individuals, households, and communities that will protections to ensure that vital assets such as be most critical to recovery. Ineffective sharing of land are not lost if people default. Specific and information on policies, programs, and plans leads focused effort will also be needed to make sure to inefficiencies in the way that individuals and that programs are accessible to people of different communities are playing their part in recovery. identities, incomes, and in different locations. Indecision has resulted from uncertainty about

71 Social Relations and Conflict

what kinds of houses people can (re)build, and critical to efficient and smooth reconstruction, people not knowing when reconstruction support especially given the overlapping mandates of gov- policy from the NRA will arrive. In addition, ernment agencies and local officials. As aid volumes confusion about roles and responsibilities have have decreased, the existing coordination mecha- led to residents being uncertain about who to nisms have become less active. More sustainable, approach for what, with local officials often medium-term platforms for communicating plans, equally confused. This has led to a continued sharing information with all relevant stakeholders, reliance on informal channels for engagement, and supporting efforts to address issues with tar- such as WCF coordinators and local political party geting are needed. representatives, with citizens communicating most effectively with actors that lack information on Effective coordination will specifically require government policy. greater geographic nuancing in aid targeting. IRM- 2 data demonstrate vast differences in experiences A key means to overcoming frustration with tar- between districts with similar levels of earthquake geting and other challenges in aid distribution is impact. Solukhumbu, for example, has been more effective communication. The high levels well relatively served by aid while assistance has of dissatisfaction with the damage assessment plummeted in Okhaldhunga. In the latter, there is process, for example, were in part caused by the evidence that some people face food security issues non-transparent, inconsistent, and confusing and that needs are great. The data in the report manner in which a series of assessments were highlight specific issues in different districts, and carried out. Citizens had little or no information the qualitative information demonstrates just how on the process and methodology used, nor on the differently recovery is experienced even between implications of the assessment for the support they wards within the same VDCs. Data from IRM can might receive. Information efforts need to better support aid providers in adjusting their responses prepare local officials to respond to queries from in ways that make them more effective, but more local residents, while also ensuring clear informa- coordination, discussion, and planning is clearly tion is spread widely across affected areas. It should needed. be anticipated that the latest NRA-led assessment will lead to changes in the classification for some Recommendation 11 – Hold focused discus- whose houses were found to be ‘fully damaged’ sions on evening out aid going to affected or ‘partially damaged’ in the second assessment. districts, including restoring aid flows to If this is not managed well, there is potential for Okhaldhunga; significant anger and inefficiency. Recommendation 12 – Revisit formal coor- Recommendation 9 – Develop a commu- dination mechanisms to generate sustain- nications strategy to reach all affected able platforms, and ensure clear roles and communities with information about responsibilities. earthquake recovery programs;

Recommendation 10 – Enhance formal ❻ Improving aid delivery will require thinking information sharing on reconstruction through targeting strategies within commu- and recovery policies between levels of nities. government, including to VDC secretaries and WCF coordinators, to ensure accurate During this round of research greater frustration information is reaching affected house- was expressed with perceived inequities in aid holds. distribution than in the early post-earthquake months. During the emergency aid distribution, handing out aid equally across and within com- ❺ Updated coordination strategies are needed munities helped to ensure social cohesion. In the for medium-term recovery efforts. recent round of IRM, complaints were shared about both targeted approaches to aid distribu- The fieldwork points to continued challenges with tion as well as equal distribution strategies. Often coordination. Information did not always flow frustrations with either targeting approach were effectively between the national, district, and VDC rooted in existing social hierarchies and tensions, levels, and between the government, NGOs, and or perceived elite control of resources. Given that foreign agencies. The DDRCs, RDCs, and other co- most new aid will likely be for housing reconstruc- ordination bodies have become less active in recent tion, it can be expected that targeted assistance will months, contributing to a reduction in information be more common. As such, it will be critical that flows between actors. Coordination efforts will be locally sensitive targeting strategies are employed

72 Aid and Recovery in Post-Earthquake Nepal

by aid providers to avoid creating resentment be- (including displaced households from that tween population groups. No matter how careful location); targeting is, resource distribution is almost always contested. Preparing local officials to engage in Recommendation 16 – Generate policy for mediating emerging disputes and preventing ten- supporting the permanent resettlement of sions from escalating will ensure that any negative displaced households unable to return to impacts from the inevitable dissatisfaction of some their villages; is mitigated. Recommendation 17 – Clarify policy for Recommendation 13 – Context-sensitive VDC and Municipality actors to mediate targeting (avoiding assumptions about and manage use of public or private land which groups are in need) is needed; this for housing displaced households in the should be based on clearly communicated near to medium term. criteria with local input;

Recommendation 14 – Expand local medi- ❽ There is a need to expand the recovery effort ation capacity for any emerging disputes beyond physical reconstruction. relating to aid. The narrative around recovery remains focused on physical recovery – the rebuilding of houses and ❼ Coherent policy for supporting displaced public infrastructure that was destroyed and dam- households should be developed. aged. This has limited the attention paid to a wide range of other issues critical for recovery. Local There is a need to address the needs of displaced needs relating to livelihoods, economic recovery, households who have moved within their wards, dealing with the psycho-social impacts of the dis- VDCs, or further away. Displaced households aster, disability, education, and health care need have been finding or building temporary shelters to be better understood and responded to. Taking on public and private land, some are migrating to a broader view of the impacts of the earthquakes, urban centers, and some small communities of and resulting needs, would help actors better pre- displaced persons are emerging. It is unclear how pare for long-term recovery. It can also help aid many of those individuals ultimately will be able providers understand how different vulnerable to return to their places of origin, and it is also groups, like children, women, and the elderly, are uncertain how long any such return might take. experiencing recovery differently and the impli- Household decisions on returning to their homes cations this has for aid programming. Qualitative are shaped by factors including perceived risks research found that some NGO programs are (including landslide risks), service disruptions providing livelihoods training to support recovery, (schools, transportation access, electricity, water but psycho-social support, women’s security, pol- sources, etc.), finances, and disruption to liveli- icy to overcome education gaps for students who hoods. There is a need to generate a clear policy have fallen behind, and other ‘soft’ issues require and processes for helping the earthquake-displaced greater attention. determine a path forward to recovery. These efforts will need at least three distinct processes. First, Recommendation 18 – Engage relevant efforts to restore services and confidence in the ministries and government actors to bet- safety of the original villages of displaced persons ter study and understand the earthquake’s are needed to enable people to move home. This impact on education, health care needs, will require geological assessments to gauge land- and inclusion; slide risk. Second, policy to support permanent resettlement for those households who will not Recommendation 19 – Increase investment return to their home villages is needed to ensure in programs to mitigate the psycho-social displacement does not lead to vulnerability and impacts of the disaster, especially for inadequate shelter. Finally, medium-term plans children. for displaced households currently living on public land are needed to avoid disputes with local communities and to ensure adequate access to services and shelter.

Recommendation 15 – Complete geological assessments in areas at risk of landslides and communicate results to communities

Independent Impacts and Recovery Monitoring Phase 2: February-March 2016 — Synthesis Report